Download PDF

Copyright is owned by the Author of the thesis. Permission is given for
a copy to be downloaded by an individual for the purpose of research and
private study only. The thesis may not be reproduced elsewhere without
the permission of the Author.
A MACROECONOMETRIC ANALYSIS OF FOREIGN AID IN
ECONOMIC GROWTH AND DEVELOPMENT IN LEAST
DEVELOPED COUNTRIES: A CASE STUDY OF THE LAO
PEOPLE’S DEMOCRATIC REPUBLIC (1978-2001)
A dissertation presented in partial fulfilment of the requirements for
the degree of
Doctor of Philosophy
in
Economics
At Massey University, Palmerston North
New Zealand
Vilaphonh XAYAVONG
2002
To the memory of my father,
Choulaphonh XAYAVONG
Page i
ABSTRACT
Despite receiving large quantities of aid, many developing countries, especially
the Least Developed Countries, have remained stagnant and became more aiddependent. This grim reality provokes vigorous debate on the effectiveness of
aid. This study re-examines the effectiveness of aid, focusing on the ongoing
debate on the interactive effect of aid and policy conditionality on sustainable
economic growth. A theoretical model of the aid-growth nexus was developed to
explain why policy conditionality attached to aid may not always promote
sustainable economic growth. Noticeable methodological weaknesses in the aid
fungibility and aid-growth models have led to the construction of two
macroeconometric models to tackle and reduce these weaknesses. The Lao
People’s Democratic Republic’s economy for the 1978-2001 period has been
used for a case study.
It is argued that the quality of policy conditionality and the recipient country’s
ability to complete specified policy conditions are the main factors determining
the effectiveness of aid. Completing the policy prescriptions contributes to a
stable aid inflow. The aid-growth nexus model developed in this study shows that
stable and moderate aid inflow boosts economic growth even when aid is
fungible. However, failure to complete the policy conditionality owing to
inadequate policy design and problems of policy mismanagement caused by
lack of state and institutional capability in the recipient country triggers an
unstable aid inflow. The model shows that unstable aid flows reduce capital
accumulation and economic growth in the recipient country. These empirical
findings reveal that policy conditionality propagated through the “adjustment
programmes” has mitigated the side effects of aid fungibility and “Dutch disease”
in the case of the Lao PDR. Preliminary success in implementing the policy
conditions in the pre-1997 period led to a stable aid inflow and contributed to
higher economic growth. This favourable circumstance, however, was impaired
by unstable aid flow in the post-1997 period. The lack of state and institutional
capacity in the Lao PDR and the inadequate policy design to deal with external
shocks triggered the instability of aid inflow, which in turn exacerbated the
negative effects of the Asian financial crisis on the Lao PDR’s economy.
Page ii
ACKNOWLEDGEMENTS
I would like to take this opportunity to express my deepest and most sincere
gratitude to Associate Professor Rukmani Gounder for her invaluable guidance,
suggestions, corrections and encouragement throughout this study. I also wish
to thank Dr James Obben for his invaluable support and supervision throughout
this study. Special thanks are due to Dr Rob Alexander for providing assistance
in applying MATHCAD to solve the simulation model.
My grateful thanks are also extended to the Ministry of Foreign Affairs and Trade
for providing me the NZODA postgraduate scholarship. I am also grateful to Jo
Donovan, Charles Chua and Sylvia Hooker for their support given to my family
during the 4 years in Palmerston North. Thanks also to Lesley Davies who
helped to proofread this thesis.
Finally, I am deeply indebted to Syrivilayphone, who has been the motivational
force in my life, and thank her for her patience, understanding and invaluable
support during the preparation of this study. My daughters and son have also
inspired me to complete my research on time.
Page iii
TABLE OF CONTENTS
ABSTRACT………………………………………………………………………………… I
ACKNOWLEDGMENT…………………………………………………………………… II
TABLE OF CONTENTS…………………………………………………………………. III
LIST OF TABLE………………………………………………………………………… VII
LIST OF FIGURES………………………………………………………………………. IX
LIST OF ABBREVIATIONS…………………………………………………….……… XII
CHAPTER 1: INTRODUCTION
1.1 FOREIGN AID AND ECONOMIC DEVELOPMENT IN DEVELOPING
COUNTRIES: AN OVERVIEW .................................................. 1
1.2 ECONOMIC IMPACT OF FOREIGN AID ON RECIPIENT COUNTRY: A
QUEST FOR GROWTH ........................................................... 6
1.3 THE ROLE OF FOREIGN AID IN ECONOMIC DEVELOPMENT OF THE
LAO PDR: AIMS AND OBJECTIVES OF THE STUDY ................. 10
1.4 CHAPTER OUTLINE ............................................................. 15
CHAPTER 2: THEORIES AND EMPIRICAL EVIDENCE OF THE
MACROECONOMIC IMPACT OF FOREIGN AID: LITERATURE
REVIEW
2.1 INTRODUCTION .................................................................. 19
2.2 AID AND POLICY IN GROWTH THEORIES ................................ 21
2.2.1 Aid-growth regression analysis in various growth
models ............................................................................ 21
2.2.2 The empirical evidence of the effectiveness of aid 30
2.3 AID AND POLICY IN DISPLACEMENT THEORIES ....................... 37
2.3.1 Aid and the savings debate ................................... 37
2.3.2 Aid fungibility and the fiscal response effects........ 40
2.3.3 Aid and the “Dutch Disease” effect ........................ 49
2.4 SUMMARY ................................................................................ 51
Page iv
CHAPTER 3: THE EFFECT OF FOREIGN AID ON ECONOMIC GROWTH AND
ECONOMIC VOLATILITY: THEORETICAL RE-EXAMINATION
3.1 INTRODUCTION .................................................................. 54
3.2 THE NEXUS BETWEEN FOREIGN AID, INCENTIVE REGIME AND
SUSTAINABLE ECONOMIC GROWTH...................................... 56
3.3 THEORETICAL MODEL OF THE AID-GROWTH NEXUS ............... 62
3.3.1 The model.............................................................. 62
3.3.2 The aid-growth nexus: simulation results .............. 66
3.4 SUMMARY ........................................................................ 73
CHAPTER 4: MACROECONOMETRIC METHODS AND DATA: FOREIGN AID
AND ECONOMIC GROWTH
4.1 INTRODUCTION .................................................................. 79
4.2 AID FUNGIBILITY MODELS .................................................... 81
4.2.1 Limitations of the existing aid fungibility models.... 81
4.2.2 A macroeconometric model of the aid fungibility ... 84
4.3 THE AID-GROWTH NEXUS MODELS ...................................... 87
4.3.1 Econometric issues of the aid-growth regression .. 88
4.3.2 A macroeconometric model of the aid-growth nexus91
4.4 ESTIMATING, TESTING AND ANALYSING MACROECONOMETRIC
MODELS............................................................................ 93
4.4.1 Estimation methods ............................................... 94
4.4.2 Goodness-of-fit testing methods............................ 98
4.4.3 Methods of analysing macroeconometric models.. 99
4.5 DATA SOURCES AND VARIABLE CONSTRUCTIONS................ 103
4.6 SUMMARY ....................................................................... 105
Page v
CHAPTER 5: AID FUNGIBILITY AND ITS EFFECTS ON INVESTMENT
5.1 INTRODUCTION ................................................................ 110
5.2 FOREIGN AID, POLICIES AND THE INTERNAL BALANCE.......... 111
5.3 ESTIMATING, VALIDATING THE MODEL AND MULTIPLIER ANALYSIS
OF AID FUNGIBILITY AND ITS IMPACT ON INVESTMENT .......... 117
5.3.1 Estimation results and model validation .............. 119
5.3.2 Empirical evidence of aid fungibility and its impact on
investments................................................................... 125
5.4 CONCLUSION .................................................................. 128
CHAPTER 6: THE EFFECT OF STABLE AID FLOW ON ECONOMIC GROWTH
6.1 INTRODUCTION ................................................................ 132
6.2 FOREIGN AID, POLICIES AND EXTERNAL BALANCES ............. 133
6.3 ESTIMATING, VALIDATING THE MODEL AND MULTIPLIER ANALYSIS
OF AID, ECONOMIC GROWTH AND POLICIES ........................ 137
6.3.1 Estimation results and model validation .............. 139
6.3.2 Empirical evidence of the interactive effect of stable
aid flow and policy conditionality on economic growth . 145
6.4 CONCLUSION................................................................... 150
Page vi
CHAPTER 7: THE EFFECT OF UNSTABLE AID FLOW ON ECONOMIC GROWTH
7.1 INTRODUCTION ................................................................ 153
7.2 STATES, INSTITUTIONS AND ECONOMIC GROWTH ................ 154
7.3 FOREIGN AID, STATE, INSTITUTIONS AND ECONOMIC
PERFORMANCES IN THE LAO PDR .................................... 162
7.3.1 Key governance Issues in the Lao PDR .............. 162
7.3.2 Foreign aid and the Lao Government’s ability to raise
investment with external viability .................................. 165
7.4 COUNTERFACTUAL ANALYSIS: THE EFFECT OF UNSTABLE AID
INFLOW ON ECONOMIC GROWTH IN THE CASE OF THE LAO PDR
...................................................................................... 171
7.4.1 The model and policy experiments ...................... 171
7.4.2 Counterfactual simulation results ........................ 175
7.5 CONCLUSION................................................................... 177
CHAPTER 8: SUMMARY AND FURTHER RESEARCH
8.1 INTRODUCTION ................................................................ 182
8.2 SUMMARY AND IMPLICATIONS ........................................... 183
8.3 FURTHER RESEARCH ....................................................... 189
REFERENCES................................................................................................... 191
Page vii
LIST OF TABLES
TABLE 1.1: DISTRIBUTION OF FOREIGN AID AMONG DEVELOPING COUNTRIES,
1970-1996 ................................................................................... 4
TABLE 2.1: AID, POLICIES, AND GROWTH RELATIONSHIPS: SELECTED RESULTS . 31
TABLE 2.2: LDCS’ FISCAL RESPONSE TO FOREIGN AID INFLOWS ...................... 47
TABLE 2.3: THE FISCAL RESPONSE EFFECT FROM FOREIGN AID INFLOWS ....... 48
TABLE 5.1: RESULTS FOR TESTING THE LONG-RUN RELATIONSHIP OF BEHAVIOURAL
EQUATIONS IN THE MACROECONOMETRIC MODEL OF AID FUNGIBILITY: FSTATISTIC .................................................................................. 119
TABLE 5.2: ESTIMATION RESULTS OF BEHAVIOURAL EQUATIONS..................... 120
TABLE 5.3: STATISTICAL TESTS FROM THE MODEL VALIDATION ....................... 124
TABLE 5.4: SHORT RUN AND LONG RUN MULTIPLIER EFFECTS OF FOREIGN AID 125
TABLE 6.1: COMPOSITION OF THE LAO PDR’S EXPORTS, 1988-1999 ............ 135
TABLE 6.2: RESULTS FOR TESTING THE LONG-RUN RELATIONSHIP OF BEHAVIOURAL
EQUATIONS IN THE MACROECONOMETRIC MODEL OF THE AID-GROWTH
NEXUS: F-STATISTIC .................................................................. 139
TABLE 6.3: ESTIMATION RESULTS OF BEHAVIOURAL EQUATIONS..................... 140
TABLE 6.4: STATISTICAL TEST OF THE MODEL VALIDATION.............................. 144
TABLE 6.5: THE SHORT RUN AND LONG RUN MULTIPLIER EFFECTS OF FOREIGN
AID FOR THE 1978-97 PERIOD. ................................................... 146
TABLE 7.1: SELECTED MACROECONOMIC INDICATORS 1978-96..................... 166
Page viii
TABLE A4.1.1 COMPARING ESTIMATED DATA WITH OTHER SOURCES ............ 107
TABLE A4.1.2: DATA USED FOR THE ESTIMATION OF THE AID FUNGIBILITY
MACROECONOMETRIC MODEL ................................................. 108
TABLE A4.1.3: DATA USED FOR THE ESTIMATION OF THE AID-GROWTH NEXUS
MACROECONOMETRIC MODEL ..................................................109
TABLE A5.1.1: SELECTED MACROECONOMIC INDICATORS: THE LAO PDR, 19852001 .................................................................................... 130
TABLE A5.1.2: SIMULATION RESULTS OF AID FUNGIBILITY MODEL: 1990-1997 . 131
TABLE A6.1.1: SELECTED MACROECONOMIC INDICATORS: THE LAO PDR, 19852001 .................................................................................... 151
TABLE A6.1.2: SIMULATION RESULTS OF THE AID-GROWTH NEXUS MODEL: 19801997............................................................................................ 152
TABLE A7.1.1: SELECTED MACROECONOMIC INDICATORS: THE LAO PDR, 19892001 .................................................................................... 180
TABLE A7.1.2: SIMULATION RESULTS OF COUNTERFACTUAL ANALYSIS: 1995-2001
.................................................................................................... 180
TABLE A7.2.1: DATA USED FOR CREATING THE BASELINE MODEL ................... 181
TABLE A7.2.2: DATA USED FOR EXPERIMENT 1............................................. 181
TABLE A7.2.3: DATA USED FOR EXPERIMENT 2............................................. 181
Page ix
LIST OF FIGURES
FIGURE 1.1: FOREIGN AID AND INVESTMENT IN THE LAO PDR, AT CONSTANT 1990
PRICES (100 MILLION KIPS) ........................................................ 14
FIGURE 3.1: THE NEXUS AMONG FOREIGN AID, INCENTIVE REGIME AND
SUSTAINABLE ECONOMIC GROWTH .............................................. 58
FIGURE 3.2: SIMULATION OF THE EFFECTS OF AN INCREASE IN A STABLE AID FLOW
ON ECONOMIC GROWTH, INVESTMENTS, SAVINGS AND CONSUMPTION
................................................................................................ 67
FIGURE 3.3: SIMULATION OF THE EFFECTS OF EXCESS IN AID INFLOW ON THE
GROWTH OF OUTPUT AND CAPITAL ACCUMULATION: THE AID-LAFFERCURVE EFFECT .......................................................................... 69
FIGURE 3.4: SIMULATION OF THE EFFECT OF AN INCREASE IN STABLE AID FLOW ON
THE GROWTH OF OUTPUT AND CAPITAL ACCUMULATION IN AN
ECONOMY WITH POOR POLICY ENVIRONMENT ............................... 71
FIGURE 3.5: SIMULATION OF THE EFFECT OF UNSTABLE AID INFLOWS ON THE
GROWTH OF OUTPUT AND CAPITAL ACCUMULATION ....................... 71
FIGURE 5.1: AID AND GOVERNMENT’S REVENUES IN THE LAO PDR, 1985-2001112
FIGURE 5.2: AID AND SAVINGS IN THE LAO PDR, 1985-2001 ........................ 113
FIGURE 5.3: FINANCIAL DEEPENING IN THE LAO PDR, 1985-2001................. 115
FIGURE 5.4: FOREIGN AID, PRIVATE INVESTMENT AND PUBLIC INVESTMENT IN THE
LAO PDR, 1985-2001 ............................................................ 116
FIGURE 5.5: SIMULATED VALUES OF GOVERNMENT CONSUMPTION SPENDING
(CGS) AND ACTUAL VALUE OF GOVERNMENT CONSUMPTION SPENDING
(CG) AT CONSTANT 1990 PRICES (100 MILLION KIPS)................ 122
FIGURE 5.6: SIMULATED VALUES OF GOVERNMENT CAPITAL EXPENDITURE (IGS)
AND ACTUAL VALUES OF GOVERNMENT CAPITAL EXPENDITURE (IG) AT
CONSTANT 1990 PRICES (100 MILLION KIPS)............................. 122
Page x
FIGURE 5.7: SIMULATED VALUES OF GOVERNMENT REVENUE (TS) AND ACTUAL
VALUES OF GOVERNMENT REVENUE (T) AT CONSTANT 1990 PRICES
(100 MILLION KIPS).................................................................. 122
FIGURE 5.8: SIMULATED VALUES OF PRIVATE INVESTMENT (IPS) AND ACTUAL
VALUES OF PRIVATE INVESTMENT (IP) AT CONSTANT 1990 PRICES
(100 MILLION KIPS).................................................................. 123
FIGURE 5.9: SIMULATED VALUES OF DISPOSABLE INCOME (YDS) AND ACTUAL
VALUES OF DISPOSABLE INCOME (YD) AT CONSTANT 1990 PRICES
(100 MILLION KIPS).................................................................. 123
FIGURE 5.10: SIMULATED VALUES OF INCOME (YS) AND ACTUAL VALUES OF
INCOME (Y) AT CONSTANT 1990 PRICES (100 MILLION KIPS) ...... 123
FIGURE 6.1: AID FLOWS AND EXPORTS IN THE LAO PDR, 1985-2001 ............ 134
FIGURE 6.2: AID FLOWS AND IMPORTS IN THE LAO PDR, 1985-2001............. 136
FIGURE 6.3: SIMULATED VALUES OF REAL OUTPUT GROWTH (GYS) AND ACTUAL
VALUES OF REAL OUTPUT GROWTH (GY) ................................... 141
FIGURE 6.4: SIMULATED VALUES OF PUBLIC INVESTMENT (IGYS) AND ACTUAL
VALUES OF PUBLIC INVESTMENT (IGY) ...................................... 142
FIGURE 6.5: SIMULATED VALUES OF PRIVATE INVESTMENT (IPYS) AND ACTUAL
VALUES OF PRIVATE INVESTMENT (IPY) ..................................... 142
FIGURE 6.6: SIMULATED VALUES OF TOTAL INVESTMENT (IYS) AND ACTUAL VALUES
OF TOTAL INVESTMENT (IY)....................................................... 142
FIGURE 6.7: SIMULATED VALUES OF CAPITAL IMPORTS (MKYS) AND ACTUAL
VALUES OF CAPITAL IMPORTS (MKY) ............................................. 143
FIGURE 6.8: SIMULATED VALUES OF TOTAL IMPORTS (MYS) AND ACTUAL VALUES
OF TOTAL IMPORTS (MY) .......................................................... 143
FIGURE 6.9: SIMULATED VALUES OF TOTAL EXPORTS (XYS) AND ACTUAL VALUES
OF TOTAL EXPORTS (XY).......................................................... 143
Page xi
FIGURE 6.10: SIMULATED VALUES OF CURRENT ACCOUNT DEFICITS (CAYS) AND
ACTUAL VALUES OF CURRENT ACCOUNT DEFICITS (CAY) ............ 144
FIGURE 7.1: THE STATE, INSTITUTIONS, AND ECONOMIC OUTCOMES............... 155
FIGURE 7.2: TRADE AND CURRENT ACCOUNT BALANCE, 1989-2001.............. 167
FIGURE 7.3: FISCAL AND MONETARY POLICY AND BUDGET DEFICITS, 1989-2001
.............................................................................................. 169
FIGURE 7.4: INFLATION AND EXCHANGE RATES VARIATION, 1985-2001.......... 170
FIGURE 7.5: FOREIGN CAPITAL INFLOWS AND CURRENT ACCOUNT BALANCE 19852001 ...................................................................................... 170
FIGURE 7.6: SIMULATION OF THE IMPACT OF VARIOUS EXPERIMENTS ON CURRENT
ACCOUNT DEFICITS .................................................................. 175
FIGURE 7.7: SIMULATION OF THE IMPACTS OF VARIOUS EXPERIMENTS ON OUTPUT
GROWTH ................................................................................. 176
Page xii
LIST OF ABBREVIATIONS
2SLS
3SLS
ADB
ADF
ARDL
AusAID
CGE
CPI
CPIA
DAC
DF
ESAF
EU
FAO
FDI
GDP
GNP
GSP
ICOR
IDA
ILS
IMF
Kips
Lao PDR
LDCs
LIML
LPRP
NZODA
ODA
OECD
OLS
PIP
PSBR
REER
RHS
SAC
SAF
SIDA
SNPA
SOEs
SURE
UK
UNCTAD
UNDP
UXO
2 Stage Least Squares
3 Stage Least Squares
Asian Development Bank
Augmented Dickey-Fuller
Autoregressive Distributed Lag
Australia Agency for International Development
Computable General Equilibrium
Consumer Price Index
Country Policy and Institution Assessment
Development Assistance Committee
Dickey-Fuller
Enhanced Structural Adjustment Facility
European Union
Food and Agriculture Organisation
Foreign Direct Investment
Gross Domestic Product
Gross National Product
General Special Preference
Incremental Capital Output Ratio
International Development Association
Indirect Least Squares
International Monetary Fund
Lao Currency Unit
Lao People’s Democratic Republic
Least Developed Countries
Limited Information Maximum Likelihood
Lao People’s Revolution Party
New Zealand Official Development Assistance
Official Development Assistance
Organisation for Economic Cooperation and Development
Ordinary Least Squares
Public Investment Program
Public Sector Borrowing Requirement
Real Effective Exchange Rate
Right Hand Side
Structural Adjustment Credit
Structural Adjustment Facility
Swedish International Development Authority
Substantial New Programme of Action
State-Owned Enterprises
Seemingly Unrelated Regression Estimation
United Kingdom
United Nations Conference on Trade and Development
United Nations Development Programme
Unexploded Ordnance
Chapter 1 __________________________________________________________________ Page 1
Chapter 1
INTRODUCTION
It is ironic and tragic that the volume of aid is declining just as
the environment for effective aid is improving. By increasing
financial assistance to poor countries with good policies and
decent institutions, we could help hundreds of millions of the
poorest people in the world to improve their lives, and those of
their children.
Dollar, 1998, quoted in the World Bank’s News Release
1.1 Foreign aid and economic development in Developing
Countries: an overview
Issues affecting the economic development of developing countries have
been on the agenda of international development cooperation for a relatively
long time.1 Since the end of the Second World War, the developing
countries have called for more favourable arrangements on international
trade and the transfer of resources from developed countries to developing
countries. The demand for this international development cooperation
gained particular momentum in the 1960s and 1970s when the number of
developing countries gaining independence rose rapidly. Within the group of
developing countries, it was soon realised that there was an even poorer
group of countries whose distinctiveness lies not only in the profound
poverty of their people but also in the weakness of their economic,
institutional and human resources, often compounded by geophysical
handicaps. The United Nations Conference on Trade and Development
(UNCTAD) classified this group of countries as the Least Developed
Countries (LDCs).2 More than any other group of developing countries, the
1
In this study “Developing countries” is used to refer to 108 countries based on the classification of the World Bank
(World Bank, 1999).
2
The criteria used to determine the LDCs are per capita GDP, share of manufacturing in total GDP, life index, economic
diversification index and population size. The following 49 countries are designated by the United Nations as least
developed: Afghanistan, Angola, Bangladesh, Benin, Bhutan, Burkina Faso, Burundi, Cambodia, Cape Verde, Central
African Republic, Chad, Comoros, Democratic Republic of Congo, Djibouti, Equatorial Guinea, Eritrea, Ethiopia,
Gambia, Guinea, Guinea Bissau, Haiti, Kiribati, The Lao PDR, Lesotho, Liberia, Madagascar, Malawi, Maldives, Mali,
Mauritania, Mozambique, Myanmar, Nepal, Niger, Rwanda, Samoa, Sao Tome and Principe, Senegal, Sierra Leone,
Solomon Islands, Somalia, Sudan, Tanzania, Togo, Tuvalu, Uganda, Vanuatu, Yemen and Zambia (this information is
available online at http://www.unctad.org/ldcs/ ).
Chapter 1 __________________________________________________________________ Page 2
LDCs are heavily dependent on external resources to expand the productive
capacity of their economies. In response to the demand for the transfer of
resources by developing countries, developed countries (donors) have
provided financial support in the hope that aid3 would enable the developing
countries (recipients) to build up their productive capacity, and in the long
run finance their investment and import requirements for self-sustaining
economic growth through normal commercial channels.4
Foreign aid is classified into two types: grant aid and loan aid. Both grant
and loan aid can be divided into bilateral and multilateral components.
Bilateral aid is administered by agencies of each donor government, such as
the New Zealand Official Development Assistance (NZODA), Australian
Agency for International Development (AusAid), Swedish International
Development Authority (SIDA), etc. Multilateral aid is funded by contributions
from developed country governments and administered by international
institutions. Some examples of multilateral institutions include the World
Bank group, the regional banks (e.g. Asian Development Bank and the InterAmerican Development Bank) and the United Nations (UN) family of
specialised agencies (e.g. the UN Development Programme (UNDP), the
United Nation Conference on Trade and Development (UNCTAD), and Food
and Agriculture Organisation (FAO)).
Foreign aid has been transferred to developing countries in the form of
project aid, commodity aid (including food aid), technical assistance, and
programme aid (balance of payments support and budget aid).5 With
accelerating globalisation and liberalisation of the world’s economy post1980, aid donors have increasingly attached policy conditionality to most of
3
In this study, “foreign aid” and “aid” are used as interchangeable terms.
4
Although donors’ objectives in providing aid is to achieve various goals such as poverty reduction, environmental
sustainability, equal income distribution and good governance, this study mainly focuses on the objective of selfsustaining growth. Also, there are many articles that write about aid motivation and describe the donor’s objectives in
providing aid as the perpetuation and extension of international capitalism and the procurement of political support from
developing countries (see for example, Hayter, 1971, 1989; Hayter and Watson, 1985; Frank, 1969, Chapter 8). The
empirical analysis of aid motivation has been undertaken by McKinlay and Little (1977, 1978a, 1978b, 1979), Maizels
and Nissanke (1984), McGillivay and Oczkowski (1991), and Gounder (1995). These areas are beyond the scope of this
study.
5
For a detailed discussions of “What is aid?” and “ Who is given aid?” see Cassen (1994, pp.2-5).
Chapter 1 __________________________________________________________________ Page 3
their aid allocation.6 To qualify for foreign aid a country must adopt economic
polices that are broadly in line with the set of policy prescriptions called the
“Washington Consensus”.7 In other words, a country is required to maintain
macroeconomic stability by controlling inflation and reducing fiscal deficits,
to expand its productive capacity by opening its economy to the rest of the
world through trade and capital account liberalisation, and to liberalise
domestic product and factor markets through privatisation and deregulation.
These sets of policy conditions have been propagated through the
stabilisation and structural adjustment programmes of the IMF and the World
Bank.8
An additional emphasis devoted to finding solutions to the economic
problems of the LDCs was established in the so-called Substantial New
Programme of Action (SNPA).9 The main objective of the SNPA was to
promote the necessary structural changes required to overcome the extreme
economic difficulties of the LDCs. Increasing food production received
special emphasis in order to augment food security and to increase
nutritional levels. The LDCs were also urged to develop their human
resources by taking the necessary steps to reduce illiteracy and to ensure a
balanced development of the various types and levels of education. Other
elements highlighted in the 2001 declaration of SNPA include strengthening
state and institutional capability to foster a people-centred policy framework
and create good governance.10
The past three decades witnessed a steady rise in aid flows from developed
countries to developing countries, despite the fact that the growth of aid
6
In this study, “policy conditionality” and “donors’ conditionality” are occasionally used as interchangeable terms.
7
Washington Consensus is the name given to the policy prescription proposed by Williamson (1990) for countries
embarking on market economic reform.
8
Apart from the stabilisation and structural adjustment measures that are imposed by the World Bank and the IMF,
“since the end of Cold War, bilateral donors have taken the lead in extending conditionality to the sphere of political
systems, introducing stipulation concerning the observance of human rights and the rule of law, and progress towards
multi-party democracy. The Bretton Woods institutions have sought to distance themselves from such overtly political
stipulation, but the enthusiastic espousal of political objectives by their major shareholders has meant that they are
inevitably drawn into this extension of attempted influence” (Killick, 1998, p.278).
9
This programme of action was declared in each of the United Nation Conferences on the Least Developed Countries.
The first and second conferences were held in Paris from 1-14 September 1981 and 3-14 September 1990, respectively.
The third conference was held in Brussels from 14-20 May 2001.
10
This information is available at http://www.unctad.org/ldcs/
Chapter 1 __________________________________________________________________ Page 4
faltered after 1992 (Lensink and White, 2000 p. 5). As indicated in Table 1.1
below, the LDCs have received more foreign aid as a percentage of Gross
Domestic Product (GDP) and aid per capita than other groups of developing
countries. Aid inflows have more than doubled in the LDCs, in that aid as a
percentage of GDP (aid per capita) increased from 10.2 percent (US$12) in
the 1970-75 period to 25.1 percent (US$29) in the 1991-96 period. Among
the group of other low-income countries aid as a percentage of GDP (aid per
capita) almost tripled from 4.7 percent (US$6) in the 1970-75 period to 12.5
percent (US$14) in the 1991-96 period. As for the rest of the group, aid as a
percentage of GDP and aid per capita in the 1991-96 period has not altered
much from the 1970-75 period, but the trend for aid flow as a percentage of
GDP has declined since the 1981-85 period.
Table 1.1: Distribution of foreign aid among Developing Countries,
1970-1996.
1970-75
LDCs
Other Low Income
Low Middle Income
Upper Middle Income
10.2
4.7
9
4.4
LDCs
Other Low Income
Low Middle Income
Upper Middle Income
12
6
1976-80
1981-85
1986-90
1991-96
Aid as % of GDP
12
13.4
18.8
25.1
6.2
6.6
10.5
12.5
12.9
10.8
10.7
7.9
4
3
3.5
1.9
Aid per capita in constant 1990 prices ($US)
15
20
29
29
7
9
14
14
9
12
13
10
10
Sources: OECD (1997) and World Bank (1997b).
By 1992, the flow of foreign aid to developing countries, calculated in 1995
prices, reached its peak of nearly US$520 billion. The cumulated amount of
actual foreign aid flow by 1996 was more than US$12 trillion (OECD, 1997).
This figure just exceeds the amount of foreign aid that is required to achieve
a growth rate of about 4 to 5 percent per year in developing countries.11
11
According to the two-gap model the well-known theoretical support of foreign aid and the growth rate of output can be
predicted by the calculation of the product between the ratio of aid to GDP and the inverse of Incremental Capital Output
Ratio (ICOR). Values of ICOR are between two to five depending on the stage of development of countries under
investigation. Chenery and Strout (1966) present a detailed discussion of the two-gap model.
Chapter 1 __________________________________________________________________ Page 5
However, developing countries have not experienced this rate of growth
since the 1980s. As Easterly points out:
In 1960-79, the median per capita growth in developing
countries was 2.5 percent. In 1980-98, the median per capita
growth of developing countries was 0.0 percent, [and] virtually
no countries outside Asia registered per capita growth at or
above the 1960-79 average of 2.5 percent.
Easterly (2000, p. 2).
In addition, it has become clear that the benefits from globalisation and
liberalisation of the world economy have been unequally distributed among
developing countries. As Nayyar indicates, 11 developing countries account
for 66 percent of the total exports from developing countries, as well as
receiving the lion’s share of foreign direct investment inflow (Nayyar, 1997
cited in Murshed, 2000, p. 2). This evidence suggests that only a few
developing countries have achieved high rates of economic growth while
others have been stagnant over the course of decades.
Whilst the 1980s were dubbed the “lost decade” for developing countries in
general and LDCs in particular, the 1990s have become, for LDCs, the
decade of increasing marginalisation, inequality, poverty and social
exclusion. The number of LDCs almost doubled from the first list compiled in
the early 1970s to 49 countries in 2001. Only Botswana has “graduated”
from the LDC group. During the 1980s, several African countries
experienced negative economic growth despite a substantial increase of aid
inflow to these countries (White, 1992a, p. 175). “A large number of
countries became more aid-dependent in the 1990s than they were in the
late 1970s” (Tsikata, 1998, p. 7).
This grim reality has raised many concerns over the effectiveness of foreign
aid. Questions such as “What is effective aid?”, “What is ineffective aid?”,
and whether aid works or not have become a substantial source of debate
among academic researchers and aid practitioners over the past few
decades. These issues are summarised in the next section.
Chapter 1 __________________________________________________________________ Page 6
1.2 Economic impact of foreign aid on recipient country: a
quest for growth
Over the past few decades, the assessment for aid effectiveness has been
approached from different ideological and methodological viewpoints. To
answer the question as to whether foreign aid contributed to economic
growth in the recipient country, aid activities have been typically assessed at
the microeconomic level. This approach has used the economic rate of
return of an individual project as a criterion for the assessment. Using this
approach, considerable success has been claimed for the effectiveness of
foreign aid. To illustrate this claim, Cassen et al. point out the assessment of
projects that have been funded by many financial institutions as follows:
... 80 percent of IDA projects achieve a rate of return of 10
percent or more. The Asian and Inter-American Bank have
concluded that 60 percent of samples of their loans fully met
their objective; 30 percent partially did so, and less than 10 per
cent were marginal or unsatisfactory. Five other major agencies
have conducted in-house reviews of large number of their
evaluations; while three of these studies remain confidential,
they all found that the great bulk of their lending had a
satisfactory rate of return…
Cassen et al., (1994, p. 8).
The above claim by Cassen et al., (1994) seems to be inconsistent with the
evidence of poor economic performance in developing countries. In addition,
many empirical studies of the macroeconomic impact of aid using data from
the 1960s onwards concluded that aid has no significant positive impact on
growth.12 Mosley called this contradiction as the micro-macro paradox and
offered three explanations for the causes of this paradox: first, inaccurate
measurement of aid effectiveness; second, fungibility of aid13 within the
public sector; and third, aid’s negative effect on investment and output in the
12
See for example, cross-country studies by Mosley et al. (1987), Reichel (1995), Boone (1994, 1996); and for specificcountry studies: of Islam (1992), Mbaku (1993).
13
Aid fungibility means that a government can increase resources through the aid inflows, to increase spending, fund
tax cuts, or reduce the fiscal deficit (reducing future tax). This would cause a negative impact of aid on growth (World
Bank, 1998).
Chapter 1 __________________________________________________________________ Page 7
private sector (Mosley, 1987, pp. 139-140). White also claims “there are
genuine theoretical reasons for not expecting the macroeconomic impact of
aid to simply be the aggregate of the micro benefits” (White, 1992a, p. 165).
He adds two more explanations to Mosley’s (1987) explanations. One is the
over-aggregation in cross-country studies of aid effectiveness. The other is
the
application
of
inconsistent
data
between
microeconomic
and
macroeconomic assessments of aid effectiveness, with the former using
economic (social) data whereas the latter uses financial (private) data
(White, 1992a, p. 164).
Furthermore, the quest for the explanation of the “micro-macro paradox”
proceeded through the impact of aid on various macroeconomic variables
affecting growth. As such, the debates have mainly hinged around the
channel through which aid contributes to growth, which suggest that:
while the microeconomic and welfare benefits of well-designed
and implemented projects may be considerable, the
macroeconomic effect - expressed in terms of overvaluation of
the exchange rate as a disincentive to exports, and continued
budget deficits as a disincentive to domestic saving and
investment - may well be negative.
White (1998, p. xvi)
The discussion of aid effectiveness in the previous paragraphs is likely to
suggest
that
the
evidence
of
foreign
aid’s
achievements
at
the
microeconomic level indicates nothing about its macroeconomic impacts,
and that both measurements are not comparable. Moreover, the discussion
is likely to suggest that foreign aid is not very useful for spurring growth in
the developing world. Indeed, this view coincided with a syndrome often
known as “aid fatigue”, especially from the early 1990s. This syndrome also
led to a downward trend of aid contributions, as the donor community has
been increasingly concerned over the effectiveness of aid. “ODA flows as a
proportion of GNP in the OECD countries have fallen from 0.4 per cent in
1975 to 0.3 per cent in 1994...[Also] aid has fallen from 66 per cent of total
resource flows to LDCs in 1986 to 40 per cent in 1994” (White, 1998, p. xv).
Chapter 1 __________________________________________________________________ Page 8
Despite the quest for the effectiveness of aid being approached from
different theoretical and methodological aspects, there is a growing
consensus that policy conditionality can no longer be regarded as a
response to a purely technical economic problem,14 and that capital
accumulation is not regarded as the only source of economic growth.15 This
view has changed the assessment of aid effectiveness by focusing on how
the economic policies and state and institutional capability of the recipient
country may contribute to growth. Much of this literature has regarded the
need for a stable macroeconomic environment, open trade regimes,
protected property rights, good quality public services, as well as political
and social stability as the factors that are important for maintaining
sustainable economic growth.
A recent study on aid effectiveness using cross-country data found that
foreign aid has a positive effect on economic growth in countries with sound
economic management and good quality public services (Burnside and
Dollar, 1997). However, aid-financed projects are always found to be
fungible (Feyzioglu et al., 1996, 1998). With respect to the impact of aid on
the private investment and trade sectors, so far “not much empirical work
has been done on the possible “Dutch disease” effects of aid, but the studies
that have been undertaken highlight the importance of an appropriate
macroeconomic policy mix to address the issues of competitiveness and the
crowding out of private investment” (Tsikata,1998).16
The above findings have provoked the donor community to reconsider the
role of foreign aid and its strategies in promoting growth in developing
countries. It was re-emphasised in the World Bank policy research report,
Assessing Aid,
…There remains a role for financial transfers from rich
countries to poor ones... in countries with sound economic
14
15
16
See for example, Haggard, et al., (1995).
See for example, Easterly (1999) and Easterly and Levine (2001).
In the aid effectiveness literature, the “Dutch disease” refers to the situation where high-levels of aid inflow may
generate undesirable effects on the recipient economy; for example, a high level of aid inflow tends to bring about real
exchange rate appreciation and hence harms a country’s international competitiveness.
Chapter 1 __________________________________________________________________ Page 9
management, foreign aid acts as a magnet and “crowds in”
private investment... In countries committed to reform, aid
increases the confidence of the private sector and supports
important public services... [thus] effective aid supports
institutional development and policy reforms are the heart of
successful development…
World Bank (1998, p. ix and p. 3).
The above discussion indicates that sound economic management and the
quality of state and institutional capability of the recipient country have
played a crucial role in the improvement of aid effectiveness. However, there
remains some dispute as to whether policy conditionality induced by the IMF
and the World Bank is sufficient to promote sustainable economic growth
(Beynon, 2001; and Lloyd et al., 2001). In fact, many Least Developed
Countries have not achieved sustainable economic growth, despite the fact
that they have made significant policy reforms (UNCTAD, 2000a). In this
context, the question that remains unanswered is why policy conditionality
attached to aid might not always promote sustainable economic growth in
Developing Countries. It is with this question that this study is predominantly
concerned.
Overall, the persistence of aid effectiveness evidence being elusive has
been attributed to an incomplete theory of the aid-growth nexus and the use
of the econometric methodology for the cross-country analysis. In addition,
the prior analyses of aid effectiveness do not include the issues of policy
conditionality relating to the quality of policy designed for aid delivery.
Recent studies of aid effectiveness also paid too little attention to the
heterogeneous
nature
of
the
developing
world.
In
particular,
the
underdeveloped nature of LDCs increases their volatility to external shocks
(i.e., natural disasters and negative spillover effects from the globalisation
and liberalisation of the world economy). Different stages of economic
development and wide cultural diversity can make state and institutional
capability vary from country to country. In this context, cross-section results
of the interactive effect of aid and policy conditionality on economic growth
should be interpreted with caution.
Chapter 1 __________________________________________________________________ Page 10
The appropriateness of country-specific studies over cross-country studies
has long been recognised (Cassen et al., 1994; Pack and Pack, 1990, 1993;
White, 1992b; and Lloyd et al., 2001). These studies indicate that there is
still much to be learnt about the impact of aid on economic growth.
Therefore, a country-specific study is needed to shed light on how economic
growth is affected by the interaction between aid and policy conditionality.
Further examination of this interaction would assist the governments of
donor and recipient countries in making aid more effective.
1.3 The role of foreign aid in economic development of the Lao
PDR: aims and objectives of the study
The Lao People’s Democratic Republic (the Lao PDR) is a landlocked
country located in the centre of the Indochina peninsula. In 1999, the
estimated population was 5 million citizens, of whom 22 percent reside in
cities and towns (World Bank, 2000). In 1975, following the end of a
protracted civil war, the Lao People’s Revolution Party (LPRP) came into
power. The Kingdom of Laos was re-named as “the Lao People’s Democratic
Republic” replacing the democratic institutional monarchy by an authoritarian
regime and a centrally planned economy.
At
the
beginning of the new regime, the country was severely
underdeveloped and much of the country’s capital stock and infrastructure
had been destroyed during the war.17 Thousands of entrepreneurs and
educated people had fled the country. The economy was dominated by
subsistence agricultural production, contributing 60 percent to GDP and
employing almost 90 percent of the labour force. The industrial sector
produced less than 10 percent of GDP and was largely composed of stateowned enterprises operating under the state planning system. The service
17
“During the 1964-1973 Vietnam War, Laos was subjected to both ground battles and aerial bombing. A total of
580,344 bombing missions were launched and more than two million tones of ordnance were dropped. Twenty years
after the end of the war, unexploded ordnance (UXO) still affects 13 (out of 17) provinces of the country, contaminating
up to 50 percent of the country’s total land area. UXO contamination threatens the livelihood and food security of the
large sections of the country’s population. It also inhibits the development of infrastructure and other services throughout
the country” (Lao Government, 2001, p.4).
Chapter 1 __________________________________________________________________ Page 11
sector was dominated by the state-owned enterprises that controlled both
foreign and domestic trades. With income per capita in 1978 estimated at
US$ 90, the country was one of the world’s least developed economies and
heavily dependent on foreign aid to meet its source requirements for
development (Otani and Pham, 1996; The Bank of the Lao PDR, various
issues).
During the 1975-85 period of pursuing economic development along socialist
lines, the government had established a highly regulated economic system.
Farm-gate prices and trade in agricultural products were administratively
determined and trade between provinces was restricted. Domestic price
controls and tight restrictions on foreign trade led to the emergence of
parallel markets for goods and foreign exchange. Private sector activities,
though discouraged, existed on a small scale. Domestic investment was
therefore dominated by the government’s capital spending on the new
establishments of state-owned enterprises and infrastructure projects, which
were still constrained by the lack of foreign exchange and low aid inflows.
During 1978-85, the ratio of investment to GDP was less than 15 percent of
GDP and the growth of real GDP was less than 2 percent annually. A
distorted incentive structure created supply shortages and a lax monetary
policy and fiscal deficit fuelled rapid inflation (Otani and Pham, 1996).
Economic stagnation was mainly attributed to the low level of foreign capital
inflow and the implementation of an inward-oriented economic strategy.18
Disappointing economic performance during the 1975-85 period and a
dramatic change in economic reform in many socialist countries led the Lao
Government to launch an economic reform programme, the “New Economic
Mechanism” in 1986.19 The main purpose of the “New Economic Mechanism”
was to achieve two transformations: first, from a centrally planned economy
18
From 1975 to the late 1980s, the Lao Government established a centrally planned economy and formed a tight
economic relationship with socialist countries. During this era, most of the Lao Government’s capital investments were
supported by economic aid from the Socialist bloc, especially the former Soviet Union. The Lao Government also
received economic aid from Western countries and other multilateral aid agencies but the amount of aid from this group
was very small.
19
It should be noted that the implementation of the “New Economic Mechanism” in the Lao PDR coincided with the
“Perestroika” in the former Soviet Union, the “Doi-Moi” in Vietnam and similar reform in many socialist countries in
Eastern Europe.
Chapter 1 __________________________________________________________________ Page 12
towards a market-oriented economy, and second, from a subsistence-based
and isolated rural economy to a manufacturing and service economy.
During the 1986-88 period of economic transition towards a market
economy,
economic
conditions
deteriorated
due
to
an
unstable
macroeconomic environment (i.e., a surge in hyperinflation and a rapid
depreciation of the exchange rate). The cause of this economic instability
was rooted in a large current account deficit and an unsustainable balanceof-payment-position. The main manifestation of this adverse effect was
attributed to the severe weakness of institutional reforms that resulted in a
lax monetary policy and lack of fiscal discipline (Bourdet, 2000, pp. 29-30).
In 1989, the Lao Government entered the stabilisation and structural
adjustment programmes under the auspices of the IMF’s Structural
Adjustment Facility (SAF) and Enhanced Structural Adjustment Facility
(ESAF), with parallel financial support provided by the Asian Development
Bank and an arrangement under the World Bank’s Structural Adjustment
Credit (SAC).
Although the carrying out of economic reform was not preceded by political
change, the Lao Government has increasingly received support from the
World Bank, IMF, Asian Development Bank and other aid donors. At the
outset of the programme, the Lao Government focused on establishing
macroeconomic stability by tightening credits and increasing budgetary
control in order to adjust external current account deficits down to the level
that could be financed by sustainable capital inflows. Many policy measures
were also taken to raise revenues, reduce and rearrange public investment
priorities and downsize the civil service, as well as applying market
instruments to manage the growth of money supply. Emphasis was placed
on the strengthening of structural production by privatising non-strategic
state-owned enterprises (SOEs) and increasing the autonomy of the
remaining SOEs. This task required the Lao Government to turn more
decisively towards reform and the establishment of various economic
institutions and legal frameworks necessary for the functioning of a market
economy.
Chapter 1 __________________________________________________________________ Page 13
To stimulate domestic resource mobilisation, the Lao Government built on
the tax system that was particularly reformed in 1988 to compensate for the
loss of government revenue previously derived mostly from the financial
surpluses of profitable state-owned enterprises. Many waves of tax reform
were carried out to broaden the tax base and many measures taken to
improve the domestic tax system and move away from a heavy reliance on
trade taxes. To develop the financial sector, major restructuring of the
banking system was undertaken. The State Bank and its 17 provincial
branches were separated into a two-tier banking system, i.e., the Central
Bank (Bank of the Lao PDR) and state-owned commercial banks. The
Central Bank was made responsible for conducting prudent monetary
policies, auctioning treasury bills, managing official international reserves,
licensing and regulating financial institutions, and establishing an effective
system of bank supervision. Entry into the banking and finance sector was
opened up to both domestic and foreign investors in order to enhance the
sector’s competitiveness.
To stimulate foreign exchange earnings, the Lao Government further
liberalised the foreign trade system. Legal and administrative measures and
foreign direct investment have been promoted under liberal and generous
policy regimes. All exports and imports other than those on specified lists
were freed from quantitative restrictions. At the end of 1992, only timber
exports remained subject to quantitative restrictions, and only imports of rice
and certain types of motor vehicles still required quantitative licensing. The
authorities removed the barriers of access to foreign exchange markets, and
exporting firms were allowed to retain and repatriate their foreign exchange
earnings. In late 1989 residents were allowed to open foreign currency
accounts with the commercial banks in order to create a more formal and
stable foreign exchange market. Efforts to raise foreign exchange earnings
from tourism were initiated in 1997 when the Lao Government launched the
“Visit Laos Year” tourist promotion.
Overall, the Lao Government’s development efforts during the last decade of
economic transformation have gone towards the establishment and building
up of the fundamental institutions and infrastructure necessary for
Chapter 1 __________________________________________________________________ Page 14
sustainable development. While the expansion in public investment is still
heavily reliant on financial aid inflows, efforts to mobilise both internal and
external resources to supplement aid inflows have been carried out under
the guidance of donors’ conditionality. Throughout the period of economic
transition up to 1997, the Lao PDR has moved decisively in the direction of
economic liberalisation and has enjoyed socio-political stability. This
economic environment contributed to a rapid increase in both public and
private investment, as illustrated by Figure 1.1 below.
Public investment
Private Investment
2001
2000
1999
1998
1997
1996
1995
1994
1993
1992
1991
1990
1989
1988
1987
1986
1985
1984
1983
1982
1981
1980
1979
1800
1600
1400
1200
1000
800
600
400
200
0
1978
100Million Kips
Figure 1.1: Foreign aid and investment in the Lao PDR, in constant 1990
prices.
Total Aid Inflow
Source: ADB (various issues), UN (various issues) and author’s estimation
Since 1989, when the Lao Government entered the stabilisation and
structural adjustment programmes, an increase in aid inflows has led to an
increase in public investment. Private investment also has risen following a
rapidly increased inflow of foreign direct investment (FDI). As a result, the
growth of real income averaged 7 percent annually from 1989 to 1997. In
this period, income per capita increased fourfold, from US$90 in 1978 to
US$400 in 1997.
However, the Asian financial crisis led to a balance-of-payment crisis and
economic volatility. This adverse effect was exacerbated by fiscal and
monetary mismanagement and in turn led to instability of aid and FDI
inflows. As Thailand is a major trading partner and the main source of FDI in
the Lao PDR, its economic downturn resulted in the Lao currency
Chapter 1 __________________________________________________________________ Page 15
depreciating by almost 90 percent against the US dollar and the inflation rate
in the Lao PDR reaching 3 digits (IMF, 2000). In 1997, the IMF and other
donors also withheld some financial aid in order to put pressure on the Lao
Government to speed up state and institutional reform, as well as to stabilise
the economy from the deteriorating macroeconomic environment. As a
result, real income growth tumbled from 6.5 percent in 1997 to about 4
percent in 1998–99 (ADB, 2000).
Assessing whether policy conditionality attached to aid promoted sustainable
economic growth in the Lao PDR requires the examination of the interaction
between aid and the incentive regimes (i.e., the state and institutional
capability to implement the policy conditions) and how this interaction
affected economic growth. In this context, this study will focus on the
potential effects of aid on economic growth through three channels. Firstly, it
examines whether foreign aid has a significant impact on tax revenue and
government spending (i.e., to address the issue of aid fungibility and its
impact on investment). Secondly, it measures the impact of stable aid inflow
on economic growth via the investment channel. Thirdly, it examines the
potential impact of unstable aid inflow on economic growth. The outcome of
this study will provide a useful insight about the channels through which
foreign aid can affect growth. It will also assist both donor and recipient
governments to address the policy implications for making foreign aid more
effective.
The period 1978 to 2001 will be studied, as this is the period for which data
are available for the Lao PDR. Most of the data used here is derived from
the annual reports of the Bank of the Lao PDR and from various issues of
the statistical yearbooks published by the Asian Development Bank and the
United Nations.
1.4 Chapter outline
This study consists of eight chapters and is organised in the following
manner. This chapter (Chapter 1) has briefly discussed the role of foreign
aid and development issues in developing countries. It has raised the
Chapter 1 __________________________________________________________________ Page 16
question as to why aid given over the past four decades has been incapable
of spurring economic growth in the developing world. The roles of foreign aid
and policy conditionality have also been highlighted as a crucial factor for
promoting sustainable economic growth in the recipient country. However,
as there remains no clear-cut explanation for the interactive effect of aid and
policy conditionality on economic growth, this raises another question as to
why policy conditionality attached to aid might not always promote
sustainable economic growth. In this context, the role of foreign aid played in
economic development of the Lao PDR has been briefly discussed. The
approaches to finding the solutions to improve aid effectiveness have also
been pointed out.
The literature review in Chapter 2 presents insights into the puzzle of why
foreign aid may not always have an positive impact on economic growth.
Many developing countries show continuing poor economic performances,
despite receiving a considerably substantial amount of aid. Thus, the
investigation focuses on the interactive effect of aid and policy conditionality
on economic growth. The relevant theories and empirical evidence of the
macroeconomic impact of foreign aid on the recipient economy are
discussed along with the two strands of aid effectiveness. The first strand of
aid effectiveness examines whether foreign aid increases economic growth
via the investment channel, while the second strand examines whether
foreign aid displaces investment via the internal balance (savingsinvestment channel) and external balance (export-import channel).
Chapter 3 expounds on the recent debate as to whether policy conditionality
attached to aid is sufficient to promote sustainable economic growth. The
analytical framework is mainly drawn from the development issues of LDCs,
including the weakness of policy conditionality designed for aid delivery. The
argument in this chapter is that donors’ conditionality and the recipient
country’s state and institutional capability are crucial factors determining the
effectiveness of aid. It is also argued that strengthening the ability to raise
investment with external viability is the key to achieving sustainable
economic growth in LDCs. In this regard, the Steger’s (2000) linear growth
model is modified to form a theoretical model of the aid-growth nexus. The
Chapter 1 __________________________________________________________________ Page 17
simulation technique is employed to demonstrate the interaction between aid
and the recipient country’s incentive regime and how this interaction may
influence economic growth and economic volatility.20
Chapter 4 presents macroeconometric methods to solve problems of the aid
fungibility and aid-growth models. Since aid has various economy-wide
effects, aid inflows tend to directly and indirectly affect both the demand side
and the supply side of an economy. The macroeconometric model of aid
fungibility is developed focusing on the potential impacts of aid on the
demand side of the Lao PDR’s economy. To take a closer look at the
potential effects of aid on the supply side of the Lao PDR’s economy, the
macroeconometric model of the aid-growth nexus is developed focusing on
the potential impacts of aid on investment and hence economic growth. In
this regard, the standard neoclassical measure of growth is modified to
integrate several aspects of the impact of capital inflow on the Lao PDR’s
economy through its role in financing domestic investment, as explained by
the two-gap model. Estimating, validating and analysing several techniques
for the macroeconometric model are discussed next. Sources of data and
methods of variable constructions employed in this study are also explained.
Chapters 5 and 6 present the effect of stable aid inflow on economic growth.
In Chapter 5, the analysis focuses on the effects of donors’ conditionality on
fiscal policy reforms. This analysis addresses the controversy issues
regarding the ongoing debate of the selective strategy for aid allocation by
examining whether aid fungibility affects investments. Therefore the
macroeconometric model of aid fungibility is employed to analyse the Lao
Government’s fiscal behaviour in response to aid inflows. Multiplier analysis
is employed to measure the short-run and long-run effects of aid on fiscal
variables (i.e., government consumption spending, public investment and
20
The incentive regime refers to the aid-receiving country’s state and institutional capability to implement the policy
conditionality.
Chapter 1 __________________________________________________________________ Page 18
government revenue). Whether aid fungibility may crowd in or crowd out
private investment is examined using the conditions of fiscal response to aid
inflow proposed by White and McGillivray (1992).
Chapter 6 supplements the analysis of the effects of aid on economic growth
carried out in Chapter 5 by focusing on the effect of stabilisation and
structural
adjustment
programmes
on
economic
growth.
The
macroeconometric model of the aid-growth nexus is employed to analyse the
effect of aid attached with conditionality on the stabilisation and structural
adjustment programmes on economic growth. Since completing donors’
conditionality is a prerequisite for the disbursement of aid, the aid multiplier
is used as a proxy to measure the short-run and long-run effects of the
interplay between aid and policy conditionality on economic growth. The
estimated multiplier values are employed to examine the Lao PDR’s ability to
raise investment with external viability.
Chapter 7 presents the effect of unstable aid inflow on economic growth. It
also examines how various types of political regimes may influence
economic growth and create unstable aid inflow. The role that the Lao PDR’s
state and institutions played in the implementation of the stabilisation and
structural adjustment programmes are analysed. It is argued that the
problem of policy mismanagement posed by the lack of state and
institutional capability is the main cause of policy slippage. This problem
coupled with the weakness of policy conditionality in turn triggered the
instability of aid inflow worsening the negative effect of the Asian financial
crisis on economic growth. The macroeconometric model developed in
Chapter 6 is modified to analyse the effect of unstable aid flows on economic
growth. The counterfactual simulation techniques are applied to disentangle
the adverse effects of unstable aid flows from the adverse effects of the
Asian financial crisis on economic growth.
Chapter 8 summarises the contributions made in this study and provides an
answer to the question of why policy conditionality attached to aid might not
always promote sustainable economic growth, with specific reference to the
case of the Lao PDR.
Chapter 2__________________________________________________________________ Page 19
Chapter 2
THEORIES AND EMPIRICAL EVIDENCE OF THE
MACROECONOMIC IMPACT OF FOREIGN AID:
LITERATURE REVIEW
...Poor countries have been held back not by a financing gap,
but by an “institutions gap” and a “policy gap”...
World Bank (1998, p. 33)
It is now widely accepted that governments complement the
market. A market economy cannot thrive, and the majority of
people cannot benefit, without wise government and effective
state institutions.
Cornia stated in Stiglitz (1998, p. V)
Aid brings a package of knowledge and finance… Aid can be
the midwife of good policies and institutions.
World Bank (1998, pp. 1-5)
2.1 Introduction
As pointed out in Chapter 1, sound economic management and the quality of
the state and institutional capability in recipient country do matter for
improving the effectiveness of foreign aid. To clarify this point, this chapter
investigates theoretical and empirical studies of aid effectiveness. Through
surveying these studies, light can be shed on the puzzle of why foreign aid
may not always have an impact on economic growth. The issues discussed
here will be empirically analysed in the case of the Lao PDR in the later
chapters.
The traditional economic justification for foreign aid is that aid will increase
growth in the recipient country. For example, foreign aid under the Marshall
Plan represented a transfer of US$13.2 billion in the 1948-52 period from the
United States of America to Europe, spurring economic recovery in that
region after the end of World War II (OECD, 1985). Theoretical support for
Chapter 2__________________________________________________________________ Page 20
this view can be tracked back to Rostow (1963) who illustrated that
developing countries in the first stage of development need foreign capital to
“kick start” their economy. In Rostow’s growth stages theory, developing
countries can then “take-off” to a stage of self-sustaining growth. Later,
Chenery and Strout (1966) put this idea into a theoretical framework, the socalled “two-gap” theory. This model incorporated the growth process of the
Harrod-Domar growth model, which considers the level of investment in
physical capital (measured by the ratio of physical capital to GDP) and the
incremental capital output ratio (ICOR) as the main driving force for output
growth. Developing countries have surplus labour but their ability to invest is
constrained by a lack of domestic savings (saving gap) and foreign
exchange availability (trade or foreign exchange gap), thus there is not
enough of these relevant resources to lead to the achievement of higher
levels of growth. In this context, the two-gap model illustrates that aid inflows
would supplement domestic savings and foreign exchange earnings one-forone. Therefore, more aid inflows will lead to higher investment and ultimately
to higher growth.
However, the theoretical basis of the two-gap model outlined above has
been challenged on various grounds. One major criticism is of the underlying
principles of the two-gap model. With an emphasis only on capital
accumulation for growth strategy, the two-gap model is too simplistic to
represent the growth process. Indeed, many aid effectiveness studies have
incorporated various growth theories to derive an analytical framework in
ascertaining the effect of foreign aid on growth. This issue is discussed in
Section 2.2. The other major criticism is the assumption of the two-gap
model which states that aid inflows will be matched by a one-for-one
increase in investment. Much of the aid effectiveness literature points out
that there are possibilities that this assumption may be incorrect (White,
1998, p. 6). Enquiry along this line has led to the development of
displacement theories, as White (1998) has dubbed them. Displacement
theories examine various links in the chain from aid to growth; this is
Chapter 2__________________________________________________________________ Page 21
discussed in Section 2.3. Finally, Section 2.4 gives a summary of theories
and empirical evidence explaining the puzzle of why aid may or may not
have an impact on growth.
2.2 Aid and policy in growth theories
The past three decades have witnessed a large number of studies on aid
effectiveness. The aid-growth nexus has been approached from different
ideological and methodological aspects. To see this, the following sections
present theoretical and empirical approaches of the aid-growth nexus model.
Firstly, Section 2.2.1 demonstrates the aid-growth regressions employed in
the various growth models. Secondly, Section 2.2.2 presents the empirical
evidence focusing on the relationships between aid, policies and growth.
2.2.1 Aid-growth regression analysis in various growth models
The earliest growth model focused on the growth of aggregate output and
resource mobilisation. The underlying analytical framework applied to
ascertain the economic-growth nexus was the Harrod-Domar growth model.
It links output growth to aggregate investment in a linear function. The rate of
output growth in the Harrod-Domar model can be captured in the production
function with capital as the sole input. This production function represents
the basic premise of developing countries, which are characterised by a
surplus of labour and the shortage of capital. Therefore, the production
function of developing countries can take the following form:
Y (t ) = f (K (t ))
(2.1)
where Y (t ) is aggregate output at time t and K (t ) is capital stock at time t. By
taking the derivative of equation (1) with respect to time (t) and dividing by Y,
this gives the growth rate of output as follows:
Chapter 2__________________________________________________________________ Page 22
Y&
1 I
= ∂K
Y
Y
∂Y
where
(2.2)
Y&
∂K
is the rate of output growth,
is the incremental capital-output
Y
∂Y
ratio (ICOR),
I
is the ratio of investment to output, and K& = I .
Y
The implication of this model is that capital accumulation is the key to
prosperity in development. This model was extended in 1966 to add a
foreign exchange constraint into the Chenery-Strout two-gap model. The
potential impact of aid on growth is simply seen as an increment to the stock
of physical capital, and can be captured in the planned investment identity,
as follows:
I = S d + A + OF
(2.3)
where S d is domestic saving, A is the inflow of aid, and OF is other source
of capital inflows.
By combining equations (2.2) and (2.3) and holding the incremental capitaloutput ratio (ICOR) constant, the rate of output growth in the two-gap model
simply depends on the accumulation of physical capital, which in turn
depends on aid inflows, domestic saving and other sources of capital
inflows. The empirical approach in the two-gap model takes the following
form:
S
Y&
A
OF
= α0 +α1 +α2 d +α 3
+ε
Y
Y
Y
Y
where
is the rate of output growth,
(2.4)
A S d OF
,
,
are respectively aid
Y
Y
Y
inflow, domestic saving and other source of capital inflows as percentage of
GDP, ε is an error term.
Chapter 2__________________________________________________________________ Page 23
Various studies published before the 1980s have been based on the above
single equation to ascertain the effectiveness of aid (for example,
Griffin,1970; Massell et al., 1972; Papanek, 1972; Gupta, 1975; and
Stoneman, 1975). Some researchers have also endogenised the right hand
side (RHS) variables of equation (2.4) to tackle the simultaneity issues (for
example, Mosley, 1980).
It should be noted that throughout the periods of aid flows there have been
several changes in development policies. Many developing countries have
had to comply with new growth strategies in order to steer their economies
into the path of sustainable growth. The change obviously makes the
empirical approach of the two-gap model (equation (2.4)) inappropriate
because it has no part that allows for capturing the change in those
development policies. Therefore, equation (2.4) may suffer from omitted
variable bias, which implies that the model would yield inconsistent estimator
of aid-growth coefficients.
Over the past few decade growth theory has undergone a radical process of
renewal and has remarkably enlarged its scope, resulting in the
development of various growth theories offering different growth strategies.
For instance, the export-led growth model (Lamfalussy, 1963) emphasises
export growth as an important factor that will encourage more investment,
bring in technical progress and increase the ability to import, thus improving
the capacity to grow. The financial liberalisation model (McKinnon, 1973;
Shaw, 1973) illustrates that financial distortion in developing countries is
caused by governments’ policies and regulations. Also, the central bank
tends to distort the real interest rate and causes credit rationing. This
distortion has an unfavourable impact on savings and investment hence
retarding economic growth. Liberalising financial markets will encourage
domestic savings in the banking system. Financial deepening1 will increase
1
Financial deepening is defined as a proportion of demand deposits to Gross Domestic Product (GDP).
Chapter 2__________________________________________________________________ Page 24
so credit rationing will be ruled out. Neoclassical and endogenous growth
theories consider productivity growth to have a leading role in the growth
process. Therefore, shifting more resources into improving human capital,
technological innovation or research and development (R&D), and
institutional quality are the keys to successful development.2 The reemergence of political economy emphasises the role of political factors in
the determination of economic growth. This literature has illustrated that
political institutions can influence the efficiency of resource allocation. Many
studies have indicated that countries are likely to sustain economic growth
where their political institutions promote socio-economic stability, create
good quality public services, offer freedom and civil liberties, provide
credible and predictable policy changes, as well as delivering stable
property rights and fair law enforcement.3 However, it is not clear which type
of political institution could provide such an environment. Furthermore, the
recent growth literature considers other factors, including inter alia the
influence of economic growth on developed countries through international
trade, as the factors that are important for the long-term growth for
developing countries (Easterly, 2000; Dollar and Kraay 2000).
To improve the estimation of the aid-growth coefficient, many studies of aid
effectiveness have incorporated a number of factors accounting for policy
and institutional quality into the model of the aid-growth nexus. The aidgrowth regression takes the following form:
Y&
A
S
OF
= α 0 + α1 + α 2 d + α 3
+ α 4Z + ε
Y
Y
Y
Y
where
is the rate of output growth,
(2.5)
A S d OF
,
,
are respectively aid
Y
Y
Y
inflow, domestic saving and other source of capital inflows as percentage of
2
For the new growth models see, for example, Barro and Sala-i-Martin (1995). See also North (1994) for the role of
institutional change influencing resource allocation.
3
See for example, Persson and Tabellini (1990), Alesina and Perotti (1995), World Development Report (1997) and
Bormer and Paldam (1998), and Gounder (2002).
Chapter 2__________________________________________________________________ Page 25
GDP, Z is a vector of control variables including the growth rate of various
factor inputs and policy variables affecting growth. Dowling and Hiemenz
(1982) add four policy variables emphasising the role of an open-trade
regime and domestic resource mobilisation in the growth process. Rana and
Dowling (1988) add the rate of export growth to the study of nine Asian
countries. In addition to domestic savings, Mosley (1987) adds growth in
literacy rates, various types of foreign capital inflows and export growth into
the aid-growth regression. Hadjimichael et al. (1995) add human capital and
various macroeconomic variables that they hypothesised to affect growth.
The macroeconomic variables used in various other studies include terms of
trade, real effective exchange rate, the rate of inflation and the size of the
budget deficit.
More recently, the research on aid effectiveness includes the possibility of
the interaction between aid and the policy index in the aid-growth
regression. The research pioneered by Burnside and Dollar (1997) focuses
on the necessity of sound policy management as conducive to sustainable
economic growth. They modify the neoclassical growth model to show how
distortionary policy may affect economic growth in developing countries. As
illustrated in neoclassical growth theory, the marginal return to investment in
poor countries is high, thus countries tend to grow faster when their capital
accumulation is rising (see, for example Romer, 1996, Chapter 1). However,
Burnside and Dollar argue that:
With a subsistence consumption constraint and imperfect
international capital markets, poor countries will tend to grow
slowly despite a high marginal return to investment. In this
context, foreign aid can accelerate growth rates in the transition
to a steady state. Furthermore, various institutional and policy
distortions can lower the return to capital and reduce
transitional growth rates.
Burnside and Dollar (1997, p. 2).
Though the role of aid in this model is similar to the two-gap model in the
sense that aid will raise investment and savings in the recipient country,
Chapter 2__________________________________________________________________ Page 26
Burnside and Dollar (1997) incorporate aid and policy variables into the
neoclassical growth framework. They point out that the aid-growth
relationship depends on the recipient country’s position on the growth path
and the level of distortion in economic policy. As such, they suggest:
…one would find a positive relationship between aid and
growth, as long as the recipient’s GDP is below the level
corresponding to its peak transitional growth rate...[also] one
would find a negative relationship between distortionary taxes
and growth.
Burnside and Dollar (1997, p. 8)
To test the interactive effect of aid and economic policy on growth, Burnside
and Dollar (1997) incorporate aid into a Barro-type growth regression.4 They
also include the interactive term for an aid variable and a policy index into
the regression. The aid variable and policy index are endogenised in order
to tackle the simultaneity issues. The analytical-framework consists of three
equations as follows:
Y&
A
AP
= β 0Y + β1Y YPC + β 2Y + β 3Y P + β 4Y
+ β 5Y Z + ε Y
Y
Y
Y
(2.6)
A
= β 0 A + β1 AYPC + β 2 A P + β 3 A Z + ε A
Y
(2.7)
A
+ β3P Z + ε P
Y
(2.8)
P = β 0 P + β1PYPC + β 2 P
where
Y&
is the growth rate of output, YPC is income per capita measuring the
Y
difference of initial income across countries,
A
is the ratio of aid inflows to
Y
GDP, P is an index measuring the distortion of macroeconomic policy,
4
For the details of the Barro growth regression see Barro (1991) and Barro and Sala-i-Martin (1995).
AP
is
Y
Chapter 2__________________________________________________________________ Page 27
the variable to capture the interactive effect of aid and the economic policy
index, and Z is a vector of control variables, including inter alia government
consumption spending, institutional quality index, political instability and
country-region variables.
So far, aid appears as a component of physical capital investment in the
model of the aid-growth nexus and is assumed to have an impact on growth
if the appropriate policies are in place. As such the level of aid flow and its
interaction with macroeconomic policy variables is included in the aid-growth
regression for testing the positive effect of aid on the recipient economy.
Nevertheless, there is a view that aid may do more harm than good to the
recipient economy when the level of aid inflow is high. This view has been
recognised in the analysis of the two-gap model, where it is also known as
the “absorptive capacity constraint” (Durbarry et al., 1998, p. 10). The
reasoning behind this view is that high aid inflows which exceed those which
can feasibly be used in profitable investment are likely to be allocated to
unprofitable investment or consumption. This is also known as the issue of
aid fungibility which will be further discussed in Section 2.3. Therefore, if a
sufficiently large amount of aid is allocated to unproductive investment, aid
will reduce productivity of investment, thus high aid inflows will have a
negative effect on growth (see for example, Griffin, 1970; Mosley et al.,
1987; and Lavy and Seheffer, 1991).
To provide a theoretical rationale, Lensink and White (1999) apply an
endogenous growth model that exhibits a negative return to aid at higher aid
levels. The underlying theoretical model is one in which the level of
technology is endogenised in the system and depends on the level of aid
flow. They show that there exists a certain value of aid flow for turning the
growth rate of output into a negative value. “The implication of this analysis
is that there may exist an aid Laffer curve: for small levels of aid, aid has a
positive effect on economic growth, while for high levels of aid, aid
negatively affects growth” (Lensink and White, 1999, p. 10). The potential
side effect of high aid inflow on the productivity of investment has also been
Chapter 2__________________________________________________________________ Page 28
emphasised in the “fiscal response” and “Dutch disease” literatures
(discussed in detail in Section 2.3). To test the possibility of an inverse
relationship between aid and growth, many researchers have included a
quadratic term of aid flow in the aid-growth regression equation, which is
basically an extension of equation (2.6) (see for example, Hadjimichael et
al., 1995; Durbarry et al., 1998; Hansen and Tarp, 1999; Lensink and White,
1999; Collier and Dollar,1999).
On the other hand, there is a view that the impact of aid on economic growth
is determined by the stability of the flow, not by the level of aid per se
(Lensink and Morrissey, 1999). As mentioned in Chapter 1, policy conditions
have been attached to aid; aid disbursement may also be subject to the
recipient country’s achieving certain criteria. For this reason, the failure of
the recipient country to fulfil such policy conditionality may relate to the
uncertainty of aid inflow, which in turn can have an adverse effect on the
level of investment and thus on economic growth. As Lensink and Morrissey
argue:
If the recipients are unsure whether they will achieve the policy
targets required to trigger the release of a tranche of aid…
uncertainty may be increased …uncertainties with respect to
[aid] inflows may render aid less effective as investors,
confronted with uncertainty, may decide to postpone or even
cancel investment decisions. Uncertainty may have similar
effects on the investment decisions and broader fiscal
behaviour of government.
Lensink and Morrissey (1999, p. 1).
Lensink and Morrissey employed three forecasting equations to capture the
expected value of aid flow. Those forecasting equations are as follows:
 A
 A
 A
  = β 0 + β1t + β 2   + β 3   + et
 Y t − 2
 Y t
 Y  t −1
 A
 A
 A
  = β 4 + β5   + β 6   + et
 Y t
 Y t −1
 Y t − 2
(2.9)
(2.10)
Chapter 2__________________________________________________________________ Page 29
 A
2
  = β 7 + β8t + β 9t + et
 Y t
where
(2.11)
A
is aid inflow as percentage of GDP, t is a time trend. Then, the
Y
uncertainty proxies or the volatility of aid flow can be obtained by calculating
the standard deviation of the residuals of those forecasting equations (i.e.
equations (2.9) to (2.11)).
Lensink and Morrissey (1999) test whether the uncertainty of aid flow affects
the relationship between aid and growth in the model of aid-growth
regression as follows:
Y&pc
Y
where
= β 0 + β 1Y pc + β 2 SECR + β 3
Y&pc
Y
I
A
+ β 4 + β 5UA + e
Y
Y
(2.12)
is the growth rate of per capita GDP, Ypc is the initial level of per
capita GDP, SECR is the initial secondary-school enrolment rate,
investment as a percentage of GDP,
I
is total
Y
A
is aid inflow as a percentage of
Y
GDP, and UA is the uncertainty proxy of aid inflow.
It should be noted that including both investment and aid variables (i.e.,
and
I
Y
A
, respectively) in the aid-growth regression is problematic because
Y
aid and investment are likely to have a strong linear relationship between
them (see equation (2.3)). In addition, the aid-growth regression could be
misspecified, as many variables that correlate with growth are excluded from
the regression (see the critique to equation (2.4)). In other words, equation
(2.12) may suffer from the presence of multicollinearlity and the omitted
variable bias, thus it could yield unreliable estimates of the aid-growth
coefficient.
Chapter 2__________________________________________________________________ Page 30
However, Lensink and Morrissey (1999) applied equation (2.12) to test the
effect of the volatility of aid flow on the relationship between aid and growth
rather than estimate the impact of aid on growth. Moreover, they estimate
equation (2.12) with and without investment in order to demonstrate the
efficiency effect of aid on growth (ibid., pp. 10-11). They also conduct a
large-scale stability analysis to test the reliability of the regression result.
The variables they include in the stability analysis are a civil liberties index,
a
political
rights
index,
political
instability
variables
and
various
macroeconomic variables (ibid. pp. 18-19).
In summary, the role of aid in the growth theories is seen as a factor input
that contributes to economic growth through a direct increment of the stock
of physical capital. Nevertheless, whether aid is effective for spurring
economic growth still depends on the policies of both donor and recipient
countries and the potential side effects of aid inflows. On the one hand, aid
may have a positive effect on economic growth if appropriate policies are in
place. This argument is based on the hypothesis that market forces in
association with the government adopting sound management may be
enough to generate economic development. On the other hand, aid may
have a negative effect on growth if a high level of aid inflows leads to the
decline in the productivity of capital investment, and if the volatility of aid
inflows leads to delaying or cancelling investment decisions. The next
section discusses the possible effects of aid on growth in the empirical
frameworks.
2.2.2 The empirical evidence of the effectiveness of aid
The empirical work on aid effectiveness has been dominated by crosscountry analysis. The selected studies of aid effectiveness shown in Table
2.1 below indicate that the impact of aid on growth is mixed. Some studies
find a statistically significant correlation between aid and growth and some
do not. For example, Papanek (1973) using data covering the 1955-65
period for a sample of 34 LDCs, obtains a significant positive impact of aid
on growth. Gupta (1975) and Stoneman (1975) obtain a similar result for a
Chapter 2__________________________________________________________________ Page 31
wider sample of LDCs and data. On the other hand, Griffin and Enos (1970)
report a negative impact of aid on growth for a sample of 32 Latin American
countries for the 1957-64 period. Voivodas (1973) uses data covering the
1956-68 period for a sample of 22 LDCs and obtains a negative impact of
aid on growth, although the aid-growth coefficient is not significant. Thus, it
is not surprising that the ambiguity of the effectiveness of aid may partly
arise from using the OLS technique for estimating the aid-growth coefficient.
Table 2.1: Aid, policies, and growth relationships: selected results
Study
Sample size and
countries case
study
Explanatory variables
Estimation
Method
Variables with significant
coefficients (sign in the
bracket)
Griffin and Enos (1970)
n=32 LA
1957-64
n=22 LDCs
1956-68
n=34 LDCs
1955-65
n=40 LDCs
1960
n=188 LDCs
1955-70
A
OLS
A(-)
A
OLS
A(-) but not significant
S, A, FDI, OF, L
OLS
A(+)
S, A, FDI, OF
OLS
A(+), OF(+)
S, A, FDI, OI
OLS
S(+), A(+),OI(-)
n=83 LDCs
1970-77
n=63 LDCs
1970-80
n=13 Asian
1968-79
S, A, I
3SLS
None
S, A, OF, GX, GL.
GX(+), GL (+)
S, A, FDI, OP, T, CG, FD
3SLS
OLS
OLS
n=71 LDCs
1980-88
n=83 LDCs
Average 1980s
N=31 S-African
1987-92
n=56 LDCs
1970-93
n=67 LDCs
1970-1988
n=58 LDCs
1970-93
n=56 LDCs
1974-93
S, A, FDI, GX, GL, PI
OLS
A(+), S(+), X(+)
S, A, GX, GDP, YPC, L,
SP
A, A2, IP, IG, Hk, N, tot,
RER, INF, BD
A, AP, YPC, EthAs, Inst,
FD, BD, INF, OP, PI, Cg
S, A, OF, GX, SA, OFA,
XA
A, A2, S, OF, tot, INF,
Bg, FD
A, A2, OP, INF, CG, FD,
YPC
2SLS
GX(+), SP(+)
GLS
All variables are
significant
AP (+), Inst (+), INF(-),
OP(+)
A(-), GX(+), SA(-)
Voivodas (1973)
Papanek (1973)
Gupta (1975)
Stoneman (1975)
Mosley (1980)
Mosley et al. (1987)
Dowling and Hiemenz
(1983)
Mosley et al. (1992)
Reichel (1995)
Hadjimichael et al.
(1995)
Burnside and Dollar
(1997)
Bowen (1998)
Durbarry et al. (1998)
Hansen and Tarp (1999)
2SLS
2SLS
2SLS
IV
A(+), S(+), FDI(+)
A(+), A2(-), S(+), OF(+),
tot(+), INF(-), BD(-)
All variables are
significant
Legend: I: total investment, IG: public investment, IP: private investment S: domestic saving, A: aid inflows, FDI: foreign
direct investment, OF: other foreign inflow, L: population or labour force, OI: other investment, OP: openness index, T:
tax revenue, CG: government consumption spending, FD: financial deepening, GX: growth rate of exports, GL: growth
rate of literacy, GDP: gross domestic product, YPC: income per capita, SP: stable prices, EthAs: interactive effect
between ethnic issues and political assassination, Inst: institutional quality, BD: budget deficit, INF: inflation, tot: term of
trade AP: interactive effect between aid and policy, Pl: policy index, LA: Latin American Countries, LDCs: less
developing countries, OLS: ordinary-least-squares, 2SLS: two-stage-least-squares, 3SLS: three-stage-least-squares,
GLS: generalised-least-squares, IV: instrumental-variables.
Sources: as shown in the column study.
Chapter 2__________________________________________________________________ Page 32
Despite many studies augmenting the aid-growth nexus model to include
policy variables and employing more sophisticated estimation techniques
during the 1980s-90s period, the results of the impact of aid on growth still
remain controversial. For example, Mosley (1980) uses data covering the
1970-77 period for 83 LDCs and employs 3SLS techniques. He finds that the
effect of aid on growth is not significant, except for the 30 poorest countries
where aid has a significant positive effect on growth. In the follow up study,
Mosley et al. (1987) introduce other explanatory variables into the
regression equation. These variables include the growth rate of exports,
growth of literacy and a variable to cover other source of financial inflows.
They apply both OLS and 3SLS technique to estimate the impact of aid on
growth. Neither method yields a statistically significant aid-growth coefficient.
They conclude that this result may relate to the issues of aid fungibility and
the crowding out of investment in the private sector. On the other hand,
Dowling and Hiemenz (1983) examine the aid-growth nexus using data
covering the 1968-79 period for 13 Asian countries, and control for a number
of policy variables such as trade, finance and government intervention. They
obtain a positive and significant impact of aid on growth.
In the last few years, the empirical studies on aid effectiveness still show no
sign of agreement. Nevertheless, a majority of aid effectiveness studies have
paid more attention to the policies which recipient governments have
followed. For example, Mosley et al. (1992) examine the role of policy
orientation in the determination of the effectiveness of aid by including the
index of policy-orientation in the aid-growth regression. Using a sample of
data covering 1980-88 for 71 LDCs, they obtain a positive coefficient of aid
and policy index variables, despite the latter being not statistically
significant. Their explanation for the insignificant statistical result of the
policy variable is due to the problem of data aggregation. They note that 55
percent of the countries in the group of “high aid, high growth” category
adopted “moderately” to “strongly outward-oriented” development strategies,
while 35 percent of the countries in the group of “low aid, low growth”
Chapter 2__________________________________________________________________ Page 33
category
adopted
“strongly
inward-oriented”
development
strategies.
Therefore, as the group of countries that adopted outward-oriented policies
is larger and the sign of policy orientation is positive, they suggest that aid
may be effective in an environment of “outward-oriented” policies.
Hadjimichael et al., (1995) investigate the impact of macroeconomic policy,
structural reforms and external factors (terms of trade, foreign aid) on
savings, investment and growth in 31 Sub-Saharan African countries for the
1987-92 period. They find that aid is effective for the group of “sustained
adjusters” who pursue stable macroeconomic policies and implement the
“adjustment programmes”, so that this group of countries is able to maximise
beneficial effects of aid and attract adequate aid inflows.
On the other hand, Bowen (1998) investigates the impact of aid on growth
through the potential interactive effect between aid and savings, investment
and export growth. His results from 2SLS estimation based on data covering
the 1970-1988 period for a sample of 67 LDCs shows that aid has a negative
impact on growth in the countries with income less than US$1000 (i.e. about
55 percent of the sample). Otherwise, aid has a positive impact on growth.
Based on this finding, he concludes that the effectiveness of aid increases
with the level of development. He also finds that aid displaced domestic
savings, which he gives as the main reason for the poor record of economic
performance in the developing countries. Boon (1994; 1996) employs data
covering the 1970-92 period for a sample of 56 LDCs to examine aid
effectiveness. He finds no significant relationship between aid and growth
and criticised recipient governments for not having appropriate economic
policies. Similarly, Burnside and Dollar (1997) examine the interactive effect
of aid and policy conditionality on growth for a sample covering 56 LDCs
over the 1970-93 period. While their results indicate a negatively
insignificant statistic of the aid-growth coefficient, they illustrated that sound
policy management is conditional for aid to have a positive effect on growth.
They conclude that aid only works when government policies are good, and
Chapter 2__________________________________________________________________ Page 34
that aid should be given to countries where governments pursue sound
policy management.
In the study by Burnside and Dollar (1997), the policy index is formed by
three variables: the level of inflation, the size of budget surplus, and the
openness to trade; the measure of institutional quality involves an
assessment of the strength of the rule of law, the quality of public services,
and the pervasiveness of corruption.5 Using these criteria, a country with
sound policy management would be one with low inflation, small fiscal
imbalances, an open trade regime, and a high score for the institutional
quality. The World Bank’s report of “Assessing Aid” cites the empirical
findings of Burnside and Dollar (1997) and concludes as follows:
Aid generally has a large effect in a good-management
environment: 1 percent of GDP in assistance translates to a
sustained increase in growth of 0.5 percentage points of GDP.
Some countries with sound management have received only
small amounts of aid and have grown at 2.2 percent per capita.
The good-management, high-aid group, however, grew much
faster [at] 3.7 percent per capita. There is no such difference
for countries with poor management. Those receiving small
amounts of aid have grown sluggishly, as have those receiving
large amounts. Introducing other variables does not change the
picture.
The World Bank (1998, p. 14).
5
This policy index has been criticized in the study by Lensink and White (2000). However, Collier and Dollar (1999)
replaced this policy index with the World Bank’s Country Policy and Institution Assessment (CPIA) score in a similar
model to that applied by Burnside and Dollar (1997), and obtain a similar result. For more details see Beynon (2001,
p.15). The CPIA has 20 different components covering four groups of policy and management measures. Those are: (1)
Macroeconomic Management which consists of general macroeconomic performance, fiscal policy, management of
external debt, macroeconomic management capacity, sustainability of structural reforms; (2) Structural Policies which
consists of trade policy, foreign exchange regime, financial stability and depth, banking sector efficiency and resource
mobilization, property rights and rule-based governance, competitive environment for the private sector, factor and
product markets, environmental policies and regulations; (3) Policies for Reducing Inequalities which consists of poverty
monitoring and analysis, pro-poor targeting and programmes, safety nets; and (4) Public Sector Management which
consists of quality of budget and public investment process, efficiency and equity of revenue mobilization, efficiency and
equity of public expenditure, accountability of the public service. Each component is scored on a 1-6 scale, and given
equal weight in computing the overall policy score.
Chapter 2__________________________________________________________________ Page 35
Although the study by Burnside and Dollar (1997) is considered pivotal to a
new generation of aid effectiveness studies, it has been criticised for its
intriguing policy implication (Hansen and Tarp, 1999). Beynon notes that:
…the original [Burnside and Dollar] research was undertaken
partly in response to critics on the right who argued that aid
was ineffective and therefore wasted, and to critics on the left
who argued that structural adjustment policies were failing…
Beynon (1999, p. 2).
The central message of the study by Burnside and Dollar (1997) is that aid
only really works when government policies are good and that aid should be
allocated to countries with good policies. This conclusion is broadly in line
with the “Washington Consensus” view on development policies (Hansen
and Tarp, 1999). According to Hansen and Tarp (1999; 2000), foreign aid
can induce growth both in countries with salutary and unfavourable policy
environments, and increased foreign aid has a positive impact on growth so
long as the ratio of aid to GDP is not excessively high or beyond a certain
threshold, the “optimal level of aid inflows”. Hadjimichael et al., (1995) and
Hansen and Tarp (1999) estimate the optimal level of aid inflows to GDP as
equivalent to 25 percent, while Durbarry et al., (1998) and Lensink and
White (1999) have estimation results which range between 40 to 50 percent
of GDP.
The crux of the ongoing dispute is the interpretation of the interactive term
(A*P) between foreign aid (A) and policy index (P) in the earlier work by
Burnside and Dollar (1997). The statistical significance of the coefficient for
this interactive term may imply that foreign aid can affect growth, but only
when policies are right. However, Lensink and White (2000, p. 6)
demonstrate that the A*P interactive term can be interpreted as either aid
can affect growth if policies are right or policies work better if supported by
aid inflow. Although the statistical robustness of this interactive term has
been re-examined in many studies, there remains no agreement as to
whether policy conditionality is a necessary condition to improve the
Chapter 2__________________________________________________________________ Page 36
effectiveness of foreign aid (Beynon, 2001, pp. 23-24).6 This may imply that
the theory of the aid-growth nexus is incomplete. Indeed, Dalgaard and
Hansen (2000, p. 7) demonstrate that the links between good policy and
foreign aid are substituted for each other in the model in which both
variables decrease the probability of social unrest, and they conclude that
“the links between aid, growth and good policy are ambiguous” on account
that “while good policies spur growth they may at the same time lead to the
decreasing effectiveness of foreign aid”. Empirically, part of the ambiguity
may be due to the methodological weaknesses of the aid-growth regression,
as the estimated aid-growth coefficient is found to be highly sensitive to the
choice of the set of control variables (Lensink and White 2000; Hansen and
Tarp, 2000).
Although there remains controversy regarding the effect of aid on growth, a
new wave of studies into aid effectiveness has acknowledged that sound
policy management is crucial for developing countries to maintain
sustainable economic growth. For example, Durbarry et al., (1998) found that
a stable macroeconomic policy environment contributes to a greater
beneficial effect of aid on growth. They also found that the growth
performance of developing countries depends on the external economic
environment, i.e. an increasing integration into the world economy through
the ongoing process of globalisation and liberalisation. Lensink and
Morrissey (1999) found that the uncertainty of aid flows has a negative
impact on investment, and hence on growth. For this reason, they suggested
that the stability of the donor-recipient relationship is crucial for enhancing
6
Other studies on aid effectiveness approach this issue by comparing economic growth between countries which did
and countries which did not adopt the “adjustment programmes”; nevertheless there remains dispute as to whether policy
conditionality works. For example, the IMF (1999a, p.2) indicates that real per capita growth of countries that enter the
IMF programmes “had caught up with that in other developing countries by the mid-1990s, …and have been rising at an
average 2.5 percent a year” for the late 1990s. However, Przeworski and Vreeland (2000, p.385 and p.395) indicate that
countries that enter the IMF programmes have a lower output growth as long as they remain in the programmes and that
countries participating in the IMF programmes have economic growth of 2.35 percent less than developing countries not
participating in the IMF programmes. Apparently, the IMF’s indication about the impact of foreign aid on growth is not
consistent with the latter studies.
Chapter 2__________________________________________________________________ Page 37
the effectiveness of aid because such stability has fostered the
implementation of appropriate economic policies.
Until now, the effectiveness-of-aid literature has demonstrated that foreign
aid contributes to economic growth through the impact of aid flow on the
accumulation of the domestic capital stock and the productivity of capital
investment. It has also emphasised the need for appropriate policy to
accommodate aid flows in order to enhance the effectiveness of aid. The
next section further explores the potential effects of aid and policy on the
accumulated capital stock and the productivity of capital investment in the
saving-investment and the export-import channels.
2.3 Aid and policy in displacement theories
Displacement theories suggest that there are possibilities that more aid
inflows may not raise investment by as much as of the value of aid inflow
and therefore an increase in aid may not lead higher rates of economic
growth. One possibility is that aid inflows may displace domestic savings
and/or “crowd out” private investment. The debate on this view has been
reproduced within the “savings debate” and the “fiscal response” literatures,
which are discussed, respectively, in Sections 2.3.1 and 2.3.2. Another
possibility concerns the impact of aid on the real exchange rate. Aid could
erode export earnings, which in turn reduces the ability to import and thus
the ability to increase investment as required. This is the subject of the
“Dutch disease” literature, which is discussed in Section 2.3.3.
2.3.1 Aid and the savings debate
The aid-savings debate was initially introduced by Griffin (1970) and Griffin
and Enos (1970). This debate focused on the hypothesis that aid inflows
may displace domestic savings. They argued that aid inflow is seen as a
complement to rather than the stimulant for income. As they indicated, the
marginal propensities to save and consume are between zero and unity,
thus aid inflows will be allocated between savings and consumption. As a
Chapter 2__________________________________________________________________ Page 38
result, savings and investment will not rise as much as the values of aid
inflows do.7 To support this argument, Griffin and Enos (1970) estimated the
aid-savings regression in the following form:
S
A
= φ 0 + φ1 + δ
Y
Y
(2.13)
Later, for a similar reason discussed in the model of the aid-growth nexus,
many researchers have incorporated more explanatory variables in the aidsavings nexus model in order to obtain unbiased estimators. The following
aid-savings regression is used to ascertain the impact of aid on savings:
S
A
Z
= φ 0 + φ1 + φ 2 + δ
Y
Y
Y
where
(2.14)
S
A
and
are the ratio of domestic saving and foreign aid to GDP,
Y
Y
respectively, and
Z
is a vector of variables affecting savings (for example
Y
the levels of exports or inflation). In this model the expected sign of the aidsavings coefficient (φ1) is negative.
However, there have been several criticisms of both the theoretical and
empirical approaches put forward in the studies by Griffin (1970) and Griffin
and Enos (1970). For example, White (1992a, 1992b and 1992c) criticised
Griffin’s model as merely a simple accounting analysis, allowing for no
feedback effect from the possibilities that aid may stimulate income and lead
to higher savings. Indeed, the analysis of Grinols and Bhagwati (1976)
incorporated the potential effect of income growth on savings, even allowing
7 This claim can be illustrated by analysing the impact of aid inflows on savings in the national accounting identity. In the
∆ I ∆S d
equilibrium position total investment identity is equivalent to I ≈ Sd + A or
≈
+ 1 , if aid inflow displaces savings,
∆A
∆A
∆S d
∆I
i.e.,
< 0 , =>
< 1 , thus investment will increase less than aid inflow.
∆A
∆A
Chapter 2__________________________________________________________________ Page 39
a negative relationship between aid and savings, which contradicted Griffin’s
view. As such, they demonstrated that aid is able to raise savings through
the dynamic effect of the increase of income on savings. Therefore, White
(1998) suggested that the greatest weakness of Griffin’s model “is probably
the fact that it holds income constant in the face of aid inflow. If aid also
affects income, then the impact on savings becomes ambiguous” (White,
1998, p. 6).
With respect to the critique of the empirical approach, there are possibilities
that the aid-savings regression equations could suffer from econometric
misspecifications
such
as
omitted
variable
bias,
simultaneity
and
multicollinearlity problems.8 Therefore, applying the aid-savings regression
equation to assess aid effectiveness is misleading and will yield ambiguous
results. Indeed, as White (1992a) indicates, “Despite some more
sophisticated treatments of the problem in recent years, there remains no
agreement as to the relationship between aid and savings” (White, 1992a, p.
189). The following paragraph highlights the empirical evidence regarding
the aid-savings nexus which have been found in various studies undertaken
since the 1970s.
The proponent of the aid-savings displacement theory estimate the aidsavings regression equations using data from both cross-country and timeseries studies and report negative values for the aid-savings coefficient
ranging between -0.73 to -0.84 (Griffin, 1970, pp. 105-106). Chenery and
Eckstein (1970) and Weisskopf (1972) also obtain a negative value for the
aid-savings coefficient. In contrast, Over (1975) finds a positive value for the
aid-savings coefficient in a similar sample to that used by Griffin (1970). In
more recent studies, the evidence on the aid-savings nexus remains
ambiguous. Snyder (1990) uses a sample of data for 50 low and middleincome countries covering the 1960-85 period and obtains a negative value
8 For a detailed discussion on the aid-savings regression critique, see White (1992b, pp.125-132).
Chapter 2__________________________________________________________________ Page 40
for the aid-savings coefficient, despite its insignificant statistical result.
However, Snyder does not rule out the possibility that there is some inverse
relationship between aid and savings because the aid coefficient was found
to have a negative relationship due to various specifications of the savings
function. Reichel (1995) uses 2SLS to estimate the model consisting of three
endogenous variables (growth, aid, domestic saving) and finds a
significantly negative coefficient for the aid-savings relationship. In a sample
of 39 sub-Saharan African countries, Hadjimichael et al., (1995) find that the
aid-savings nexus varied depending on the growth performance and the
degree to which adjustment efforts were sustained. In the group of
“sustained adjusters”, foreign aid appears to have stimulated domestic
savings, whereas in a group of countries with negative per capita growth rate
and protracted economic imbalances, foreign aid has a negative impact on
savings.
It is not surprising that the empirical evidence on the aid-savings nexus is
mixed and ambiguous. These results reflect the theoretical weakness of the
aid-savings displacement hypothesis. Although the analysis of the aidsavings nexus does not provide clear-cut information about how aid may
affect growth through this channel, further enquiry in this area has continued
in more detail. Attention is directed towards aid fungibility issues which
involve the investigation of the relationships among aid inflows, fiscal policy
variables and investment. This is discussed below.
2.3.2 Aid fungibility and the fiscal response effects
As argued in the aid-savings displacement theory, Griffin (1970) indicated
that recipient countries tend to substitute aid inflow for domestic resources.
As most of the aid goes to support public expenses, recipient governments
may reduce their tax efforts. If this is the response recipient governments
pursue to accommodate aid inflows, it could create an unfavourable
environment for encouraging domestic savings and private investment.
Indeed, by reducing tax revenues, governments could face chronic budget
Chapter 2__________________________________________________________________ Page 41
deficit problems, as government spending could rise to accommodate the
escalation of aid inflows. Sooner or later the recipient governments may not
be able to avoid the need to print money and/or to raise public sector
borrowing requirement (PSBR) to finance their budget deficits. These moves
could affect private savings (investment) through the negative effects of high
inflation (high interest rates), which may occur due to the growth of highpowered money (the increase in PSBR will bid up the interest rate). Besides,
the government may change the composition of its expenditure towards
unproductive investment and/or consumption (i.e., the issue of aid
fungibility).9 The influence of foreign aid on public investment may thus be
unproductive and not promote economic growth.10
There are two main approaches to the estimation of aid fungibility
coefficients. The most popular of the aid fungibility models, the “fiscal
response” model, was first developed by Heller (1975). This model was
further extended by Mosley et al. (1987), Gang and Khan (1991) and
Franco-Rodriguez et al. (1998). In this approach, the government minimises
a loss function subject to expenditure constraints to obtain empirical
frameworks in the form of structural or reduced-form equations. The aid
fungibility coefficients can be obtained by estimating the fiscal behavioural
coefficients from such an equation system. Another aid fungibility model was
developed by Pack and Pack (1990; 1993), in which the aid fungibility
coefficient can be obtained by regressing government spending on a range
of variables. Although this approach is an atheorectical model, it can tackle
the essence of the bureaucratic decision-making process (Pack and Pack,
1990, p. 189). Other researchers applying similar approaches to that of Pack
and Pack (1990; 1993) are, for example, Khilji and Zampelli (1991; 1994),
Feyzioglu et al. (1998) and Swaroop et al. (2000). Each of these studies
9
The issue of aid fungibility is another strand in aid-effectiveness studies. Aid fungibility has been widely explored at
both aggregate and sectoral levels. However, this study focuses only on evaluating aid fungibility at the aggregate level.
Some recent studies of aid fungibility are Pack and Pack (1991), Khiji and Zampelli (1991) and Mitsuhashi (1996).
10
See Griffin (1970, pp. 103-107) for a detailed discussion on this issue.
Chapter 2__________________________________________________________________ Page 42
derives the empirical framework from a utility maximising problem. Aid is said
to be fungible at the aggregate level if the following conditions are satisfied:
∆IG
<1
∆A
(2.15)
∆CG
>0
∆A
(2.16)
∆T
<0
∆A
(2.17)
where CG, IG, T and A are government consumption spending, public
investment, taxation and aid inflows, respectively.
Equation (2.15) indicates that aid intended to be used for public investment
does not increase public investment to the extent of the value of aid inflow.
This implies that the recipient government is able to avoid donor attempts to
target aid, and some of released resources that are available due to an
increase in aid inflow can be used to increase consumption (equation (2.16))
or to fund tax cuts (equation (2.17)). To estimate the aid fungibility
coefficient, the first approach employs a structural model which is
represented by a system of equations as follows:
DI t = f1 ( DI t* , Tt , At )
(2.18)
IDI t = f 2 (CGt* , IDI t* , Tt , At )
(2.19)
CGt = f 3 (CGt* , IDI t* , Tt , At )
(2.20)
Tt = f 4 (Tt * , CGt* , CGt , DI t , At )
(2.21)
Equations (2.18) through to (2.21) consist of fiscal-choice variables. These
variables are: direct public investment (DI), indirect public investment (IDI)
(here IG= DI+ IDI), government consumption spending (CG), government
Chapter 2__________________________________________________________________ Page 43
revenue (T) and total aid (A) which can be classified into grant aid and loan
aid. An asterisk (*) denotes the fiscal-target variables which are defined in
equations (2.22) through to (2.25) as follows:
DI t* = f 5 (Yt −1 , IPt )
(2.22)
IDI t* = f 6 ( Enrt , Yt , GYt )
(2.23)
CGt* = f 7 (CG )
(2.24)
Tt* = f 8 ( Yt , M t −1 )
(2.25)
where Y is real income and GY is growth of real income, IP is private
investment, Enr is primary school enrolments and M is imports. Estimation of
the target variables in equations (2.22) through to (2.25) is crucial for the
subsequent estimation of the aid fungibility coefficients in equations (2.18)
through to (2.21).
Other approaches to estimate the aid fungibility coefficient are represented
by equations (2.26) through to (2.29):
Tt = f 9 ( At , Z t )
(2.26)
CGt = f10 ( At , Z t )
(2.27)
IGi ,t = f 11 ( Ai ,t , A≠i ,t , Z t )
(2.28)
n
CGt + ∑ IGi ,t = Tt + At + PSBRt
(2.29)
s =1
where T, CG, IG, and A, are defined as before, Ai,t is aid allocation to sector i
at time t, , A≠i,t is aid allocation to other sectors at time t, PSBR is public sector
borrowing requirements and Z is a vector of control variables. Pack and Pack
(1990, 1993) use GDP as the control variable, while Feyzioglu et al. (1998)
use the infant mortality rate, average years of schooling in the labour force,
average ratio of a neighbouring country’s military expenditure to GDP, and
Chapter 2__________________________________________________________________ Page 44
ratio of agricultural output to GDP. Equation (2.28) is employed to capture
sectoral-aid fungibility coefficients of the public investment (i.e., the
∆IGi ,t
coefficients of
∆Ai ,t
). The aid fungibility coefficient of the public investment
at the aggregate level (i.e., the coefficient of
∆IG
) can be obtained by
∆A
summing up the sectoral aid fungibility coefficients.
Although the fiscal response model provides useful information about the
effectiveness of aid, it implicitly indicates that aid fungibility could affect
economic growth through affecting the marginal productivity of aid. To solve
this puzzle, Mosley et al. (1987) extended Heller’s framework to incorporate
a production function. This analytical framework highlights the relationship
among aid, growth and private investment. The impact of aid on growth can
be derived from the simplified equations noted by White (1992a), as follows:
U =γ0 −
γ1
γ
2
2
(
G − G * ) − 2 (T − T * )
2
2
(2.30)
G = γ 3 A + γ 4T
(2.31)
G * = γ 5 IP
(2.32)
T* = 0
(2.33)
IP = γ 6 A
(2.34)
Y = Y (KP , KG , L )
(2.35)
where U is the government’s utility in the form of a welfare-loss-function, G is
government spending (G=IG+CG), G* is the government spending target, T
is taxation, T* is the taxation target, A is foreign aid, IP is private investment,
Y is production, KP and KG are respectively, fixed capital stock of private and
government sectors, and L is the labour force.
Chapter 2__________________________________________________________________ Page 45
In this model, equation (2.30) implies that the welfare loss function depends
on how a government’s budget is allocated. The government can maximise
the value of social welfare so long as its actual and target budget levels are
the same (i.e., at G = G * and T = T * ). Thus the value of the welfare loss
function is positive, (U = γ 0 >0), indicating that the social welfare function is
maximised, and any allocations that deviate from the target will lower the
value of social welfare (i.e., U =γ1 <γ 0 ). Equation (2.31) is the government’s
budget constraint, assumed to depend only on aid inflows and taxation
revenue. To simplify the model, the taxation target is set equal to zero in
equation (2.33). Equation (2.32) captures the influence of government
spending on private investment, as governments can set their spending
targets to promote private investment. In other words, governments can set
their spending targets to encourage private investment. Equation (2.34)
captures the direct impact of aid on private investment. On the supply side,
production is driven by the stock of private and public capital and the labour
force (equation (2.35)).
By minimising the welfare loss subject to a budget constraint, the model
yields a set of equations that determine the coefficient of governmental
behaviour. The substitution of the reduced forms of IG (obtained from first
order condition) and IP into the growth function gives the solution for the
impact of aid on growth, as follows:
  ∂Y µ

∂Y&  ∂Y
  ∂Y µ
=
γ 6  + 
γ 3  + 
γ 5γ 6 
∂A  ∂KP   ∂KG 1 + µ   ∂KG 1 + µ

where µ =
γ2
γ 1 (γ 4 )
2
(2.36)
> 0.
Equation (2.36) illustrates that aid affects growth through three channels.
The first term captures the impact of aid on growth through private
 ∂Y 
investment (γ 6 ) and the marginal productivity of private capital 
 . The
 ∂KP 
Chapter 2__________________________________________________________________ Page 46
second term captures the impact of aid on growth through public investment
(γ 3 )
 ∂Y 
and the marginal productivity of public capital 
 . The last term
 ∂KG 
captures the impact of aid on growth through the interactive effect of public
and private investment (γ 5γ 6 ) and the marginal productivity of public capital
 ∂Y 
.

 ∂KG 
Despite the coefficient of government behaviour (µ) entering into the last two
terms of equation (2.36), this does not give a clear-cut view as to whether
the issue of aid fungibility can be considered as a sufficient condition to
solve the puzzle of why aid may or may not have an impact on growth. As
White (1992a) argued, the fall in tax revenue due to the increase in aid
inflows might not be observed. This is because aid may raise current income
and hence increase tax collection. He further elaborated that the impact of
the low level of public savings on private savings and investments perhaps
depends on how governments finance their spending. Therefore, the most
striking implication of equation (2.36) is that the impact of aid on growth
mainly depends on the impact of aid on the private sector and the impact of
aid fungibility on the productivity of public and private capitals, which in turn
depend on the government’s sound policy management. In this context,
White and McGillivray (1992) applied a simple economic model to examine
the relationship between aid and private investment with various sets of
fiscal responses. The model consists of two equations as follows:
I p = I p ( r ) and I p′ < 0
(2.37)
where I p is private investment, which is assumed to have an inverse
relationship with the real interest rate (r ) .
PSBR = G − ( T + A)
(2.38)
where PSBR, the public sector borrowing requirement, is equal to the
government’s budget deficit.
Chapter 2__________________________________________________________________ Page 47
Table 2.2 below provides a summary of the sets of fiscal response to aid
inflows under various conditions which may explain why aid fungibility may
or may not have impact on private investment, and hence on economic
growth.
Table 2.2: LDC fiscal response to foreign aid inflows
Case
1
2
3
4
5
6
7
8
9
10
Conditions
∆T ∆G
=
=0
∆A ∆A
∆PSBR
⇒
= −1
∆A
∆T
∆G
= 0, 0 <
<1
∆A
∆A
∆PSBR
⇒
<0
∆A
∆T
∆G
= 0,
=1
∆A
∆A
∆PSBR
⇒
=0
∆A
∆T
∆G
= 0,
>1
∆A
∆A
∆PSBR
⇒
>0
∆A
∆T
∆G
= 0,
<0
∆A
∆A
∆PSBR
⇒
< −1
∆A
∆T
∆G
1<
< 0,
=0
∆A
∆A
∆PSBR
⇒
<0
∆A
∆T
∆G
< 0, 0 <
<1
∆A
∆A
∆ PSBR
∆G ∆T
⇒
< 0 if
−
<1
∆A
∆A ∆A
dT
dG
< 0, 0 <
<1
dA
dA
dPSBR
dG dT
⇒
> 0 if
−
>1
dA
dA dA
∆T
∆G
< 0,
>0
∆A
∆A
∆ PSBR
∆G ∆T
⇒
= 0 if
−
=1
∆A
∆A ∆A
∆T
∆G
> 0,
>1
∆A
∆A
∆PSBR
⇒
>0
∆A
Description
Government uses aid to finance planned expenditure,
taking opportunity to reduce PSBR (i.e., reducing future
government expenditure); present value of reduction equals
the value of aid inflow.
Effect on Ip
Crowding In
Weaker version of Case 1 in which (1-dG/dA) is used to
reduce PSBR, with remainder used to increase government
expenditure.
Crowding In
No fungibility with all aid used to finance increased
expenditure.
None
Government increases expenditure by more than the value
of aid inflows (as in, for example, counterpart funding
requirement); PSBR increases.
Crowding
Out
Government reduces expenditure in response to aid inflows
on belief that aid expenditure is more efficient than
domestically financed expenditure.
Crowding In
Aid inflows used to both offset PSBR and reduce taxes.
Crowding In
Aid inflows partly used to finance increased expenditure
and partly to reduce taxes, net result being a decrease in
PSBR.
Crowding In
Aid inflows partly used to finance increased expenditure
and partly to reduce taxes, net result being an increase in
PSBR
Crowding
Out
Aid inflows partly used to increase expenditure and partly to
reduce taxes, net result being no change in PSBR
None
Government increases expenditure by more than the value
of aid inflows; taxation revenue increases due to an income
multiplier effect.
Crowding
Out
Source: White and McGillivray (1992)
To evaluate the fiscal response effect from the aid inflow or the impact of aid
fungibility on growth, the coefficients of the impact of aid on tax, government
spending and the public sector-borrowing requirement are needed. Table 2.3
below provides empirical evidence of the fiscal response effect and adding
Chapter 2__________________________________________________________________ Page 48
aid fungibility variables to the information provided by White and McGillivray
(1992).
Table 2.3: The Fiscal Response Effect from Foreign Aid Inflows
∆G
∆A
∆T
∆A
n=11African
1960s-70s
0.96
-0.4
0.36
Cashell & Craig.
(1990)
n=46 LDC
1975-80s
3.60
0.00
2.60
McGuire
(1987)
Israel
-0.52
0.00
Pack and Pack
(1991)
Indonesia
1966-86
1.58
Gang & Khan
(1991)
India
1961-84
Khilji & Zampelli
(1991)
Pakistan
Study
Data
Heller
(1975)
∆PSBR
∆A
Is aid fungible
in aggregated
level?
Yes
Effect on private
investment
Crowding Out
(case 4)
-1.52
Yes, soft-loan
& grant
are likely
No
0.29
0.29
No
Crowding Out
(case 10)
-0.55
0.00
-1.55
No
Crowding In
(case 5)
0.26
-0.74
0.00
Yes
None
(case 9)
Crowding Out
(case 8)
Crowding In
(case 5)
Sources: as shown.
As can be seen from Table 2.3, the existence of aid fungibility is not always
associated with the negative impact of aid on private investment. This is not
a surprising result because the issue of aid fungibility is not a sufficient
condition to indicate that aid may or may not have a negative impact on
growth. Nevertheless, in the case of the aid fungibility crowding out private
investment, the negative impact of aid on the private sector is due to the
pressure on recipient governments to increase PSBR, which in turn could bid
up the interest rate and lower investment demand. Moreover, there is a view
that the negative impact of aid on the private sector could be channelled
through the high level of aid inflows exerting upward pressure on the
domestic price level, especially on non-tradable prices. As Mosley et al.,
claimed:
the transfer of aid money raises the prices of some goods,
depresses the prices of some others and hence has side
effects on the private sector.
Mosley et al. (1987, p. 617).
Chapter 2__________________________________________________________________ Page 49
This claim, however, could indicate that aid inflows might have dual impacts
(positive and negative) on the recipient economy. On the one hand, aid
inflows exert upward pressure on non-tradable prices that may encourage
producers to produce more non-tradable goods. On the other hand, the
increase in non-tradable prices might cause the appreciation of the real
exchange rate if non-tradable prices rise faster than tradable prices. The
latter is the issue known as “Dutch disease” effect, which is discussed in the
following section.
2.3.3 Aid and the “Dutch Disease” effect
In the aid effectiveness literature, “Dutch disease” is used to refer to the
situation where high levels of aid inflow may generate undesirable effects on
the economy (Edwards and Van Wijnberger, 1989). “Dutch disease” arises
when the high level of aid inflows brings about real exchange rate
appreciation and creates booming sector (non-tradable sectors) at the cost
of recession in other sectors (tradable sectors). The symptoms of “Dutch
disease” can be observed once the increase of aid inflows draws resources
away from tradable into non-tradable sectors. As a result, tradable
production declines and hence threatens export performance (Corden and
Neary, 1982). It is obvious that the effect of “Dutch disease” will erode the
recipient’s export earnings and hence the ability to import. Therefore, more
aid inflows which may cause the “Dutch disease”, will not be matched by a
one-for-one increase in investment (White, 1998, pp. 6-7).11
As “Dutch disease” arises due to the high level of aid inflow creating a
booming sector in the economy, it is important to analyse the level of aid
inflow that may cause the “Dutch disease” effect. White (1992a) indicates
that the “Dutch disease” effect happens in both the developed and
11 This claim can be illustrated by analysing the impact of aid inflows on export earnings in the national accounting
∆M ∆X
identity. In the equilibrium position the import identity is equivalent to M ≈ X + A or
≈
+ 1 , if aid inflows
∆A ∆A
∆M
∆X
displaces export earnings, i.e.
< 0 =>
< 1 , imports increase less than aid inflows.
∆A
∆A
Chapter 2__________________________________________________________________ Page 50
developing world, however, sector booms in the developed world, unlike
those in the developing world, are not the result of high capital inflow. The
“Dutch disease” syndrome arises when the booming sector accounts for a 6
to 15 percent share of GDP in developed countries, but in developing
countries, the level of aid inflow to GDP varies from more than 5 percent to
over 20 percent. Therefore, White suggests that there is a high potential that
aid may bring in the “Dutch disease” effect (White, 1992a, pp. 166-67). In
the past decade, studies of the “Dutch disease” effect have been
approached from different methodologies. These range from the application
of the single regression equation of aid-real exchange rate nexus models to
the establishment of macroeconometric models or computable general
equilibrium models (CGE). The appropriateness of an approach depends on
the extent of aid inflows and the considerable economy-wide effects of aid
on the recipient economies. Empirical evidence regarding the “Dutch
disease” effect is highlighted below.
Van Wijnbergen (1986) applies a single regression equation to estimate the
aid-real exchange rate nexus model for African countries. He finds a
significantly negative relationship between aid and the real exchange rate in
four out of six African countries. He also demonstrates that the effect of the
“aid boom” permanently lowers the total productivity in the export sector.
Despite the real exchange rate being allowed to depreciate after the effect of
the “aid boom”, productivity does not return to the level before the “aid
boom”. Nevertheless, he argues that if capital markets were perfect, there
should have been no problem after the effect of the “aid boom” as the private
sector can re-borrow and re-invest after the economic recovery from this
effect.
Using the CGE model, Weisman (1990) investigates the impact of aid
inflows to Papua New Guinea. He finds that aid inflows increased
government spending, which in turn increased the prices of non-traded
goods and services. Producers responded to the increase in prices of nontraded goods by increasing supply in this sector and shifting resources from
Chapter 2__________________________________________________________________ Page 51
the production of traded goods. Therefore, aid inflows brought about the
“Dutch disease” effect that threatened the export earnings of Papua New
Guinea. Collier and Gunning (1992) also apply the CGE model to examine
“Dutch disease” effects in African economies. They find that aid supported
government spending that raised aggregate demand and exerted upward
pressure on the prices of non-tradable sectors. As a result of the booming of
non-tradable sectors, labour and capital were drawn away from the tradable
sector. They illustrated that devaluation does reduce this inverse effect on
tradable sectors.
In summary, the “Dutch disease” literature regards the high level of aid
inflow as the potential source of side effects on the recipient economy. For
this reason, aid may not have a positive impact on growth if high levels of aid
inflow make tradable sectors less competitive in the world market through
the appreciation of the real exchange rate and the lowering of export
earnings. Eventually, the “Dutch disease” effect will lower the ability to
import, invest and grow.
2.4 Summary
This chapter has surveyed theoretical and empirical aspects of the quest for
aid effectiveness. This survey has focused on the interrelationship between
aid and economic policies in the growth and displacement theories. It has
also examined the empirical framework employed to estimate the aid-growth
and aid fungibility coefficients in various models. The response to why aid
may not always have an impact on economic growth can be summarised
from the theoretical and empirical points of view as follows:
Theoretically, aid has economy-wide effects on the recipient country. Aid is
said to be effective if an increase in aid raises savings, investment and
export earnings. However, that aid may not always be effective has been
attributed to the problems of the recipient country’s policy mismanagement.
The problems that can make aid less effective are as follows:
Chapter 2__________________________________________________________________ Page 52
§
Aid is fungible. This means that the excess aid inflow may be used to
invest in low productivity sectors and/or to increase government
consumption spending and/or to fund tax reduction. The former two
would make aid less effective while the latter would deter domestic
savings and investments via upward pressure on prices and interest
rates.
§
Aid inflows bring in a source of “Dutch disease”. This means that high
levels of aid inflow cause the overvaluation of exchange rates. Aid
would undermine the external competitiveness of the recipient country
thus crowding out investments in the traded goods sector and export
earnings.
§
The failure of the recipient country to accomplish donors’
conditionality leads to the uncertainty of aid inflow. This could lead
both public and private sectors in the recipient country to postpone or
even cancel investment decisions.
Empirical evidence of aid effectiveness has been mainly produced from
cross-country analyses. Studies published before the late 1990s found
mixed evidence regarding the aid-growth relationship and aid fungibility
issues. However, aid shows a positive impact on growth once the models
incorporate the side effects of aid as a quadratic term into the aid-growth
regression. Indeed, many studies conclude that aid contributes to economic
growth so long as the level of aid inflow is stable and not excessive. The
estimate of the optimal level of aid ranges from 25 to 50 percent of GDP.
The finding implies that aid does work although aid is fungible. Therefore,
there is no need to resolve the micro-macro paradox puzzle.
While a positive effect of aid on economic growth has been found, the
interactive effects between aid and policy conditionality remain ambiguous.
For example, there has been an increasing consensus that countries with
stable and open economic environments (e.g. low inflation, small fiscal
imbalances, a free trade regime), give prominence to law and order as well
as having good quality public services are likely to achieve sustainable
economic growth. However, these studies do not provide a clear-cut
explanation on the interactive effect of aid and the recipient country’s
policies on sustainable economic growth. In fact, some studies of aid
Chapter 2__________________________________________________________________ Page 53
effectiveness demonstrate that aid has also been found to make a positive
contribution
to
economic
growth
even
in
countries
hampered
by
unfavourable policy environments. The persistence of this puzzle is
attributed to the theoretical and methodological weaknesses of the aidgrowth relationship. This also gives rise to ongoing debate as to whether
policy conditionality is a necessary and/or sufficient condition for aid to have
a positive impact on economic growth. To clarify this puzzle, the quality of
policy conditionality and the recipient country’s incentive regime becomes
vital in examining the effects of aid on sustainable economic growth. In other
words, how the recipient country pursues economic policies to raise
investment with external viability is a crucial factor to ensure sustainable
economic growth. As such, the enquiry into the potential effects of aid on
economic growth and economic volatility along the path of maintaining
external viability will be further explored in Chapter 3.
Chapter 3__________________________________________________________________ Page 54
Chapter 3
THE EFFECTS OF FOREIGN AID ON ECONOMIC GROWTH
AND ECONOMIC VOLATILITY: THEORETICAL REEXAMINATION
The economic history of the world is replete with recessions
and depressions. From the bursting of the British South Sea
Bubble and the French Mississippi Bubble in 1720 (which at
least one economic historian claims delayed the industrial
revolution by 50 years) to the industrial depressions of the
1870s and 1930s, to the Latin American middle income debt
crisis, African low income debt crisis, ex-Communist output
collapse, and East Asian financial crisis…it is a wonder that
anyone in the world has economic security.
Easterly et al. (2000, p. 1)
To a great extent, the obstacles to the successful participation
of vulnerable developing economies in the international system
are rooted in the causes of their underdevelopment and poor
economic performance. Nevertheless, the new rules of the
game and the international economic environment prevalent
since 1980 following accelerated globalization, leaves them
vulnerable in novel ways.
Murshed (2000, p. 1)
3.1 Introduction
The assessment of aid effectiveness discussed in Chapter 2 reveals that
foreign aid has contributed to economic growth, although there remains
some dispute as to whether policy conditionality is sufficient to support
economic growth. It may be plausible to claim that foreign aid has a positive
impact on economic growth because in past decades aid has been mainly
allocated to investments in physical and human capital, factors presumed to
be the fundamental driving force for economic growth.1 But why have the
majority of developing countries not achieved sustained economic growth,
1
According to data from the Geographical Distribution of Financial flow to Developing Countries on CD-ROM (OECD,
2001), about 60 to 70 percent of total aid flow during the 1973 to 1999 period has been allocated to investments in
physical and human capital.
Chapter 3__________________________________________________________________ Page 55
despite the receipt of substantial amounts of foreign aid over the past
decades?
As indicated in Chapter 1, the median per capita growth of developing
countries was zero during 1980 to 1998. In fact, developing countries have
not only experienced a low record of economic growth but also a high record
of economic volatility (Easterly et al., 2000, p. 4). Recent research by Gavin
and Hausmann (1998) shows that economic volatility correlates negatively
with economic growth and that a volatile macroeconomic environment leads
to a significantly lower rate of investment, undermines educational
attainment, harms the distribution of income and increases poverty. Apart
from terms of trade fluctuation and other external shocks, macroeconomic
policies (i.e., fiscal and monetary policy and the exchange rate regime) and
political instability are the most important determinants of economic volatility
(Gavin and Hausmann, 1998, pp. 97-98). Although policy index, institutional
quality and political instability variables have been included in the
assessment of aid effectiveness, none of the previous studies explicitly
focused on the interrelationship among these factors. To solve this puzzle, a
clear understanding is required of how the interaction between aid and the
recipient country’s incentive regimes (i.e. the state and institutional capability
to implement the policy conditions) may affect economic growth and
economic volatility. In this context, this chapter introduces the theory of the
aid-growth nexus focusing on development issues of the Least Developed
Countries (LDCs).
The rest of this chapter is organised as follows. Section 3.2 sketches the
analytical framework for explaining the interactive effect of aid and the
recipient country’s incentive regime on economic growth and economic
volatility. Section 3.3 presents the theoretical relationship within the aidgrowth models that will be used to demonstrate the impact of aid flow on
savings and capital formation, as well as the interactive effect of aid and the
recipient country’s incentive regimes on economic growth and economic
volatility. Finally, Section 3.4 summarises and draws some implications for
Chapter 3__________________________________________________________________ Page 56
the empirical investigation of the aid-growth nexus that will be undertaken in
the case of the Lao PDR’s economy.
3.2 The nexus between foreign aid, the incentive regime and
sustainable economic growth
Although the aid-growth nexus has long been a source of vigorous debate in
aid effectiveness literature, there are few theoretical models linking aid and
growth. As discussed in Chapter 2, the theoretical nexus between aid and
growth is mainly based on the Harrod-Domar and two-gap models, in which
aid is seen as a driving force for economic growth via its contribution to
capital accumulation in the recipient country. More than twenty years passed
before further theoretical aid-growth models were presented by Mosley et al.
(1987), Boone (1996), White (1992a), Burnside and Dollar (1997), Lensink
and White (1999) and Dalgaard and Hansen (2000). The first two studies
link the impact of aid fungibility and economic growth. While Mosley et al.
(1987) demonstrate that aid fungibility could affect economic growth via the
marginal productivity of public and private capital; Boone (1996) found that
aid has no effect on long run growth because aid is used as substitute for
consumption. Burnside and Dollar (1997) demonstrate that good policy
improves the effectiveness of aid, while Lensink and White (1999)
demonstrate that excessive aid inflow over the optimal level has a negative
effect on economic growth. White (1992a) also shows in detail the
macroeconomic impact of aid-growth relationship. In contrast to the first
three studies, Dalgaard and Hansen (2000) demonstrate that aid fungibility
is not the main problem for the effectiveness of aid and that the interactive
effect of aid and good policy on growth is ambiguous.
To re-examine the aid-growth nexus, the following paragraph outlines the
policy conditionality designed for aid delivery, as well as various factors that
are believed to be obstacles to obtaining sustained economic growth in
Least Developed Countries.
Chapter 3__________________________________________________________________ Page 57
Like other developing countries, LDCs have undertaken the “adjustment
programmes”2 in order to promote sustained economic growth and improve
the well being of their population. Since 1988, “thirty three out of the 48
LDCs have engaged in SAF or ESAF programmes” (UNCTAD, 2000a, p.
103).3 The policy conditionality induced by the World Bank and the IMF
through the “adjustment programmes” is broadly in line with policy
prescriptions under the Washington Consensus, and has been designed to
support economic growth with external viability, i.e. external current account
deficits could be financed by normal and sustainable capital inflows
(UNCTAD, 2000a, p. 102). This adjustment strategy was based on the view
that poor economic performance in LDCs during 1980s and 1990s was
attributed to the persistent external debt burden and the lack of external
finance. The theoretical justification for this adjustment strategy can be
tracked back to the “two-gap” model of Chenery and Strout (1966).
According
to
the
“two-gap”
model,
developing
countries
face
the
fundamental constraint of lacking enough domestic savings and foreign
exchange to meet the desired level of domestic investment. Consequently,
such countries are limited in their ability to achieve a target growth rate and
hence rely on external finance, including foreign aid, to fulfil their destiny.
For self-sustaining economic growth, they have to mobilise domestic
resources to finance investment and reduce their heavy reliance on external
finance. In addition, they have to raise foreign exchange earnings to finance
imported intermediate and capital goods as well as repay their debts.
Whether they can break these financial constraints and fulfil their destiny
2
3
“Adjustment programmes” are hereafter referred to as the stabilisation and structural adjustment programmes.
The Washington Consensus suggests 10 policy prescriptions for countries embarking on a market economic reform.
These are: (1) keep fiscal deficit to minimum, (2) make public expenditure priority toward education, health and
infrastructure development, (3) reform tax system by lowering marginal tax rates and broadening the tax base, (4)
liberalise interest rate to discourage capital flight and increase savings, (5) adopt a competitive exchange rate to bolster
exports, (6) liberalise trade and minimise tariffs for intermediate goods needed to produce exports, (7) encourage foreign
direct investment, (8) privatise the state-owned enterprises , (9) deregulate the economy to allow enterprises entry and
exit market freely, (10) secure property rights via law and order enforcement in order to encourage saving and
accumulating wealth. For more details see Williamson (1990).
Chapter 3__________________________________________________________________ Page 58
depends on their ability to increase investment with external viability.4 Given
that
“adjustment
programmes”
are
designed
to
overcome
the
abovementioned financial constraints, the potential interaction between
foreign aid and the recipient country’s incentive regime may affect economic
growth. This is illustrated in Figure 3.1 below.
Figure 3.1: The nexus among foreign aid, incentive regime and
sustainable economic growth
Aid donors
Recipient
government
Policy conditionality attached to aid
(Donors’ conditionality)
Incentive Regime
Ability to raise foreign exchange
earnings to finance imports and to
repay foreign debts
Ability to mobilize domestic resource to
finance investments and reduce
dependence on external finance
Ability to increase investment with
external viability
Three engines of growth
Economic
growth
Economic
volatility
Short run
Growth
Long run
Sustainable
economic growth
4
The view on demand side driven economic growth suggests that output growth is subject to balance of payment
constraints. Thirlwall (1980) illustrates that the balance-of-payments constrained growth rate depends on capital inflows
and trade performance, and that “in the long run a country cannot grow faster than the rate of growth of output consistent
with the balance-of-payments equilibrium on current account” (Thirlwall, 1980. p. 251). For an up-to-date survey of the
literature, see McCombie and Thirlwall (1997). On the supply-side approach, at least during the early stages of economic
growth and development, it is necessary to accumulate capital from foreign capital. As foreign capital increases, it could
raise the productivity of capital through the introduction of modern technology and management and by raising the skill
base of the labour force. For the role of capital accumulation on economic growth and development in underdeveloped
countries, see for example Nurkse (1953), Lewis (1955) and Arndt (1987). Moreover, to achieve sustainable economic
growth it is required to establish a virtuous circle among savings, exports and investments. In this context, an increase in
exports not only raises foreign exchange for imports and investments, but also provides markets for goods which would
not otherwise be produced, or produced only to meet domestic demand. As investments increase, accelerated economic
growth is expected, which in turn increasingly provides additional resources for capital accumulation though an increase
in the ability to save (UNCTAD, 1999, Part II, and UNCTAD, 2000b).
Chapter 3__________________________________________________________________ Page 59
The flow chart shows the nexus between foreign aid, the recipient country’s
incentive regime and three strategies to achieve sustainable economic
growth. The double-dotted arrow shows the indirect nexus between foreign
aid and investment via policies designed to enhance both internal and
external resource mobilisations. The bold-dash arrow shows the direct
foreign aid-investment nexus, as foreign aid is mainly used as budget
support, which directly raises public investment. Policy conditions are
officially set out in the Letter of Intent for financial support to the IMF (i.e.,
the policy conditions designed for aid delivery). In this letter, recipient
governments spell out the details and time frame in which to implement an
agreed three-year programme of policy commitments that often combine with
the conditionality of other aid donors such as the World Bank and Asian
Development Bank and consist of hard-core conditionality and pro forma
elements. Completing the hard-core conditionality is a prerequisite for the
approval of, or continued access to, foreign aid.5
Based on the above discussion, a country would need to raise investment
with external viability in order to achieve sustainable economic growth.
Success of this goal also depends on establishing a virtuous circle between
investment, exports and savings (i.e., conducting a prudent policy to keep
three engines of growth running). In this context, aid attached to the
“adjustment programmes” was expected to help developing countries to
take-off and reach the stage of self-sustaining growth in due time. However,
from past development experience it has become clear that many developing
countries, in particular LDCs, remained stagnant and become more aid
dependent than they were previously. This poor economic performance is
attributed to the weakness of the policy conditionality designed to deal with
external shocks.
As explained earlier, the “adjustment programmes” have been designed to
5
For more details on the connection between structural policy designs and the monitoring of the aid delivery system see
IMF (2001).
Chapter 3__________________________________________________________________ Page 60
support economic growth at the level that external current account deficits
can be financed by normal capital inflow. In practice, the policies designed to
accomplish such an objective have often overlooked the underdeveloped
nature of the LDCs. As indicated in the Least Developed Countries Report
(UNCTAD, 2000a, p. 116), “terms-of-trade deterioration and shortfalls in
external financing were often considered by IMF staff to be risks ex ante, but
contingency measures and adjusters were not built into the programme”. And
under the stabilising policy prescription, “any financing shortfall would have
to be offset fully and immediately by a tightening of policies or a contraction
of imports” (Mecagni, 1999, p. 236). However, it has been acknowledged
that stabilising policy prescriptions, which seek to reduce aggregate demand
as quickly as possible, often create political difficulty and even lead to
nationalistic reaction (see for example, Haggard, et al., 1995).
Poor economic performance in the LDCs is also attributable to policy
mismanagement and the lack of enforcement of the rule of law, which often
relates to the lack of state and institutional capability to create good
governance.6 These problems are also compounded by many factors such
as: (i) the extreme poverty in the LDCs that makes it very difficult for the
mobilisation of domestic resources to be realised; (ii) the structural
weakness of LDC economies characterised by lack of infrastructure
(telecommunication, roads, electricity, etc); (iii) a vulnerable indigenous
private sector which makes supply side capacity very rigid and heavily
dependent on foreign direct investment (FDI); and (iv) the complex
international trade barriers and lack of markets for key exports making it
difficult for LDCs to maintain a current account balance.7 These are the
biggest problems affecting the LDCs’ ability to raise investment with external
viability.
6
The problems of policy mismanagement and the lack of enforcement of the rule of law posed by the lack of state and
institutional capability that may have an influence on economic growth and economic volatility will be discussed in detail
in Chapter 7.
7
See Murshed (2000) for a detailed discussion of the obstacles to the successful participation of the LDCs in the
international system.
Chapter 3__________________________________________________________________ Page 61
The abovementioned obstacles are not only a fundamental source of LDCs
exposure to the negative effects of domestic and external shocks but also
the causes of policy slippage (i.e., the interruptions that often result in LDCs
failure to accomplish the policy conditionality they agreed to). As the
disbursement of foreign aid is subject to the prior action and ability of the
LDCs to fulfil certain policy conditions, it is quite often the case that the IMF
withholds financial supports when efforts to implement the “adjustment
programmes” are interrupted. Mecagni (1999) classifies the causes of
programme interruption into three episodes: (1) no major policy slippage but
there is disagreement over policy formulation in response to external shocks,
(2) political disruptions that are serious enough to prevent meaningful
negotiations or call into question the continuing authority of current
government, and (3) policy slippage arising from fiscal issues and structural
constraints which are often linked to four circumstances: (i) external shocks,
(ii) natural disasters, (iii) social unrest related to structural adjustment
effects, and (iv) democratisation or pre-electoral climate.
The Least Developed Countries Report (UNCTAD, 2000a, Table 26)
indicates that a substantial proportion of programme interruptions occurred
due to the tension between policy conditionality and the recipient country’s
national sovereignty (i.e., policy ownership issues), in which 14 out of 34
interruption episodes were attributable to policy disagreements between
recipient governments and the Fund’s staff. This is not surprising; as it has
long been acknowledged that conditionality is regarded not only as a
technique to tackle economic issues but also to handle political issues (see
for example Haggard et al., 1995). Five episodes of programme interruptions
are related to political turmoil and social unrest. The incidence of the
interruptions is either strongly or weakly linked to the effects of the
“adjustment programmes” themselves. The rest of the interruptions are
related to external shocks such as terms-of-trade deterioration and natural
disasters (UNCTAD, 2000a, Table 26).
Chapter 3__________________________________________________________________ Page 62
When an interruption occurs, it disrupts capital accumulation in the recipient
country. As Sachs et al., (1999, p. 7) point out, suspension of the IMF’s aid
disbursement can deny the recipient government access to other sources of
concessional finance from other major creditors, including the World Bank
and bilateral donors. It is often the case that the interruption is followed by
difficult negotiations to restart the programme. There can thus be a
prolonged shortage of foreign exchange resulting in an inability to maintain
external viability. This in turn worsens the budget situation, creating a
balance-of-payments crisis, economic volatility, and instability of aid inflow.
3.3 A theoretical model of the aid-growth nexus
The issues indicated above are further discussed in this section including
the theoretical model of the aid-growth nexus. A linear growth model
developed by Steger (2000) is extended to include the potential effects of
aid inflows. The steps of developing the model and its properties are
explained in Section 3.3.1. The potential of interactive effects between aid
and the recipient country’s incentive regime on growth are examined in
Section 3.3.2.
3.3.1 The model
The linear growth model developed by Steger (2000) is based on a closed
economy
framework
to
explain
economic
growth
with
subsistence
consumption. This indicates that a country only relies on internal resources
to finance domestic investments. This assumption is not appropriate for
LDCs. The model utilized in this study differs from Steger (2000) by
employing an open economy framework to explain the aid-growth nexus. As
discussed in Section 3.2, LDCs rely heavily on both internal and external
resources to expand their ability to raise domestic investment. To
incorporate the external resource constraints into the model, a capital
accumulation equation is modified to include the net per capita capital inflow
variable. In this context, the evolution of per capita capital stock is equal to
the sum of domestic savings and the net capital inflows minus the
Chapter 3__________________________________________________________________ Page 63
depreciation of per capita capital stock, and thus can be expressed as
follows:
k&( t ) = s( t ) + f ( t ) − ( δ + n )k( t )
(3.1)
where s(t) is per capita savings, f(t) is the net per capita capital inflow, n is
constant rate of population growth and δ+n is the depreciation rate of per
capita capital stock.
As illustrated in Figure 3.1, LDCs require the ability to mobilise domestic
resources and raise foreign exchange earnings in order to achieve
sustainable economic growth. In equation (3.1), s(t) and f(t) capture the
LDCs’ ability to raise investment from internal and external resources,
respectively. The LDCs’ ability to maintain economic growth with external
viability is also captured by f(t). This is because f(t) represents net per capita
capital inflow at the sustainable level of the external account. Any changes
that affect the flow of external capital (i.e., unstable aid inflow) will affect the
LDCs’ ability to raise investment. Therefore, equation (3.1) captures the
LDCs’ ability to raise investment with external viability.
Based on equation (3.1), it is further assumed that net per capita capital
inflow includes only foreign aid inflows. It is also assumed for now that net
per capita aid inflow grows at a constant rate ( µ ) . Therefore, the amount of
net per capita aid inflow takes the following form:
f (t ) = F0 e µt
(3.2)
where f(t) and ( µ ) are defined as above and F0 is the initial stock of capital.
The model is also extended to include a modification of Steger’s (2000)
distortion index in the production function, which reflects the level of
structural distortion in the economy created by poor policies and institutional
rigidities in the LDCs. It is further assumed that the only factors of production
are capital and labour and that technology exhibits constant returns to scale.
Although this kind of technology is simple, it can capture a change in
Chapter 3__________________________________________________________________ Page 64
productivity that can be derived from the improvement of technical and
management skills gained from learning-by-doing and knowledge spillovers
(Barro and Sala-i-Martin, 1995, p. 140). Besides, this production function
has possible implications for desirable government policy (ibid., p. 140). It is
therefore appropriate to apply constant returns to scale to capture the
positive effects of technology transfers and changes to the liberal policy
regime because of aid inflows. For the sake of simplicity, the size of the
labour force and population are assumed to be equal so that per capita
output net of distortion can be expressed in the following form:
y(t ) = (1 − π )Φk (t )
(3.3)
where y( t ) is per capita output, k ( t ) is per capita capital stock, π is the
distortion index, and Φ is total productivity.
Now, a benevolent social planner with dictatorial powers will choose to
maximise the utility of the representative household subject to the economy’s
resource constraints, thereby allowing the complete model to read as
follows:
∞
Max ∫
c( t )
0
[c( t ) − c ]1−θ
1 −θ
− 1 −( ρ − n )t
e
dt
subject to resource constraints k&( t ) = s( t ) + f ( t ) − ( δ + n )k ( t )
k ( 0 ) = k 0 , k(t) ≥ 0 for ∀ t, and c ≤ c(t) ≤ Φk(t)
(3.4)
where c is the subsistence level of per capita consumption, θ is a constant
preference coefficient and ρ is the individual time preference rate. Following
the standard way to solve the preceding dynamic problem, it can be
expressed as the optimal path of per capita consumption and capital, as
follows:8
8
See Appendix 3.1 for a detailed mathematical solution to the model.
Chapter 3__________________________________________________________________ Page 65
c( t ) = c + [c( 0 ) − c ]eθ
[
][
−1
k ( t ) = k + k ( 0 ) − k eθ
[(1−π )Φ −δ − ρ ]t
−1
[(1−π )Φ −δ − ρ ]t
(3.5)
+ F1 e µt
]
(3.6)
where k is the per capita subsistence capital stock and its value is defined
as follows: k ≡
c
( 1 − π )Φ − δ − n
In the case of no foreign aid inflow (i.e., F1=0), the model has a similar
property to the linear growth model with subsistence consumption developed
by Steger (2000), in which the economy is caught in the poverty trap when
both per capita consumption and capital stock are equal to their subsistence
level and the net marginal product of capital is relatively lower than the time
preference. Intuitively, at the subsistence level of income individuals do not
have the ability to save or the ability to raise investment (ibid.). Escaping the
poverty trap in the world without foreign aid can only be achieved when the
net marginal product of capital exceeds the time preference rate and
consumption exceeds the subsistence level (ibid.). In other words, whenever
individuals are willing to save and have the ability to save and invest, the
economy will take off. Both per capita consumption and capital stock will
grow and reach their asymptotic balance-growth equilibrium at a certain
time.
In the case where a country receives foreign aid, policy conditionality
attached to aid will raise the ability to invest. The growth rate of per capita
consumption and capital stock will approach asymptotic balanced-growth
equilibrium at the following rate:
lim
c&( t )
→ θ −1 [(1 − π )Φ − δ − ρ ]
t →∞ c( t )
(3.7)
k&( t )
→ θ −1 [(1 − π )Φ − δ − ρ ]
t →∞ k ( t )
(3.8)
lim
Whenever the economy moves on the transitional path from the subsistence
level of income to the self-sustained balanced-growth equilibrium, the
Chapter 3__________________________________________________________________ Page 66
savings rate is increased along with per capita income and approaches its
asymptotic balanced-growth equilibrium at the following rate:
 [(1 − π )Φ − δ − ρ ]  k ( t ) − k 
sy( t ) = 


θ [(1 − π )Φ − δ − n] k ( t ) 
lim s y ( t ) →
t →∞
[(1 − π )Φ − δ − ρ ]
θ [(1 − π )Φ − δ − n ]
(3.9)
(3.10)
It should be noted that the economy would no longer need foreign aid when
it achieves its asymptotic balanced-growth equilibrium. At this stage the
property of economic growth is similar to Steger’s (2000) linear growth
model. As Steger demonstrated, economic growth is driven by the savings
rate (similar to equation (3.9)), which in turn is determined by two terms. The
first term represents the willingness to save which is determined by the
preference and technology coefficients (ibid.). The second term represents
the ability to save or invest which is measured by the relative distance
between the starting value of the capital stock and its subsistence level
(ibid.). The closer the ability to save and invest is to zero (i.e.,
a ≡ k( 0 ) − k / k( 0 ) ≈0), the longer the time required for the economy to take
off (ibid.).
3.3.2 The aid-growth nexus: simulation results
To analyse the aid-growth nexus, a simulation technique is employed. The
coefficient values used here are similar to those employed by Steger
(2000).9 The economy is assumed to start at the point where the ability to
save and invest are close to zero, aid inflow is assumed to raise the ability to
invest from a=0.009 up to a=0.09, and the per capita capital stock is
assumed to grow at the rate of 1 percent per year. The simulation results are
displayed in Figure 3-2 below.
9
Coefficients used for the simulation are: Φ=0.1, θ=3, δ=0.02, ρ=0.01, n=0.025,
MATHCAD is employed to perform the simulation analysis.
c =2, F1=0.009, µ=0.01, a=0.009, and
Chapter 3__________________________________________________________________ Page 67
Panel (a) of Figure 3.2 indicates that an increase in aid raises the growth
rate of per capita income by shifting up the transitional growth path of per
capita income, and that an increase in aid by 1 percent per year shortens the
time required for the economy to achieve its balance-growth equilibrium
value.
Besides, as illustrated in Panel (b), an increase in aid shifts up the
transitional path of savings rate to a relatively higher path than in an
economy without aid inflows. It is therefore obvious that an increase in aid
inflow gives rise to an increase in per capita income growth and domestic
savings.
Figure 3.2: Simulation of the effects of an increase in a stable aid flow
on economic growth, investments, savings and consumption
(a)
(b)
Chapter 3__________________________________________________________________ Page 68
(c)
Note: the solid line represents the case of no aid inflow while the dotted line represents the impact of aid inflows.
Panel (c) of Figure 3.2 shows the effects of aid on consumption and
investment which are in line with the model used by Dalgaard and Hansen
(2000). They demonstrate that aid fungibility is not the main problem for the
effectiveness of aid because an increase in aid inflow gives rise to an
increase in the expected return on investment, which in turn is a reason for
consumption growth along with increased aid inflows (Dalgaard and Hansen,
2000, pp. 5-6). On the transitional path both consumption and capital pursue
an unbounded growth rate. It is therefore not surprising to observe an
increase in consumption along with an increase in aid inflows in the
economy. This substitution effect is varied along the transitional path.
As illustrated in Panel (c) of Figure 3.2, an increase in aid inflow causes a
downward shift of the transitional path of the consumption-capital stock ratio.
Initially, the ratio declines sharply. This indicates that the growth path of
capital stock is rising faster than the growth path of consumption. In other
words, the substitution of investment for consumption is small at the
beginning of transitional path; therefore aid fungibility is not important.
When the economy reaches its optimal level, it is expected that consumption
will increase at a similar rate to the increase in the growth rate of capital
stock, in which the consumption-capital stock ratio changes at a constant
rate when it approaches its balanced-growth equilibrium. This is so because
the need to improve standard of living can be achieved only by keeping
investment growth at the same level as consumption growth.
Chapter 3__________________________________________________________________ Page 69
Figure 3.3 below shows the negative effect of a high level of aid inflow on
economic growth. The model simulation revealed a prediction that is
consistent with the model of the “Aid Laffer Curve” used by Lensink and
White (1999), in which increased aid inflow beyond a certain level gives rise
to a negative effect on economic growth.
Figure 3.3: Simulation of the effects of excess aid inflow on the growth
of output and capital accumulation: the Aid Laffer Curve effect
(a)
(b)
Note: In Panel (a) of Figure 3.3, the dash represents the increased path of aid with the accumulative rate of 1.5 percent
per year, while the dotted line represent with the accumulative rate of 1 percent per year. In Panel (b) of Figure 3.3, the
solid line represent the case of no aid inflows while the dot line represents the impact of aid inflows
As illustrated in Panel (a) of Figure 3.3, the growth rate of aid inflow is
assumed to have increased by 50 percent (from the dotted line to the
Chapter 3__________________________________________________________________ Page 70
dashed line). The increased aid boosts the economy to overshoot the steady
state equilibrium growth rate of per capita capital stock.10 As illustrated in
Panel (b) of Figure 3.3, after the overshooting point, the economy travels
along the optimal growth path and approaches the asymptotic-balanced
growth equilibrium with a negative growth rate of per capita capital stock so
that the increased aid inflow causes a negative rate of economic growth on
average.
The analysis of the interaction between aid and the recipient country’s
incentive regime on economic growth and economic volatility is carried out
by employing two different simulation approaches. Firstly, the simulation
process takes into account a favourable impact of aid on economic growth
that is impaired by the distortions of economic structure created by poor
policies and institutional rigidities. This assumption is based on the view that
the problems of structural distortion have resulted from the implementation of
an inward-oriented development strategy, and the lack of state and
institutional capability to enforce the rule of law. In this context, poor policies
and institutional rigidities render the supply side of the economy less
responsive to market signals. Therefore, poor policies and the lack of state
and institutional capability could create high transaction costs; thus lowering
the country’s total productivity and leading to a decline in economic growth
(North, 1990).
The simulation of this effect is illustrated in Figure 3.4, in which the
distortions created by the poor policies and the lack of state and institutional
capacities lower the transitional growth rate of per capita capital stock and
economic growth in an economy with a stable aid inflow.
10
The growth rate of aid inflow causing the economy to overshoot its steady state can vary and depend on coefficients
employed in the simulation model.
Chapter 3__________________________________________________________________ Page 71
Figure 3.4: Simulation of the effect of an increase in stable aid flow on
the growth of output and capital accumulation in an economy with poor
policy environment
Note: the solid line represent the case of a stable aid inflow while the dotted line represents the impact of economic
distortion created by the recipient government’s poor policies and deficient institutions. The distortion index is assumed
to increase from π =0 to π=0.02.
Secondly, the simulation process takes into account a favourable impact of
aid on economic growth that is impaired by instability in aid inflows. As
explained earlier, the cause of instability in aid inflows is attributable to the
shortcomings of donors’ conditionality coupled with the recipient country’s
policy mismanagement posed by the lack of state and institutional capability.
This in turn creates economic volatility and hence lowers the process of
capital accumulation. To capture the effect of unstable aid inflows on the
growth of output and capital accumulation, the flow of aid is assumed to
fluctuate over time. Here the growth rate of aid flow is assumed to grow at
the rate of µ(t)=cos (t). The simulation result of unstable aid flow is displayed
in panel (a) of Figure 3.5 below.
Figure 3.5: Simulation of the effect of unstable aid inflows on the
growth of output and capital accumulation
(a)
Chapter 3__________________________________________________________________ Page 72
(b)
Note: In panel (b) of Figure 3-5, the solid line represents the case of a stable aid inflow while the dotted line represents
the impact of unstable aid inflows.
Panel (a) of Figure 3.5 shows that aid flows are interrupted twice a 10-year
period. The impact of this unstable aid flow is illustrated in Panel (b), in
which the instability of aid inflow reduces the growth rate of per capita capital
stock and hence per capita income. As financial aid is mainly given for
budget support, the decline in aid inflow by itself lowers public investment,
which in turn, has a negative effect on the indigenous private sectors reliant
on government contracts. Also, as discussed in Chapter 2, when investors
are faced the uncertainty of aid inflows, they may delay or cancel investment
decisions.
From the above analysis, it can be said that the extent to which aid may
have a negative effect on economic growth depends on whether the
interplay between aid and the LDCs’ incentive regimes creates either
structural distortions or economic volatility. Given UNCTAD’s (2000a) report
that LDCs have moved decisively in the direction of economic liberalisation
(i.e., removed structural distortions resulting from the implementation of the
inward-oriented development strategy), it is not hard to claim that LDCs’ poor
economic performances are rooted in the policy mismanagement posed by
the lack of state and institutional capability, and the shortcomings of the
policy conditions designed for aid delivery.
Chapter 3__________________________________________________________________ Page 73
3.4 Summary
This chapter has provided a theory of the aid-growth nexus, in which the
interactive effect of aid and the recipient country’s incentive regime on
sustainable economic growth has been established on the premise of
financial constraints long recognised in the two-gap model. Given that aid
donors have increasingly attached policy conditionality, the model developed
in this chapter approaches this issue by focusing on how the interactive
effects of aid and policy conditionality may influence the recipient country’s
ability to raise investment from internal and external sources. Whether
foreign aid may or may not contribute to sustainable economic growth is
assessed through the recipient country’s ability to raise investment with
external viability.
To solve the remaining dispute as to whether policy conditionality is
sufficient to promote sustainable economic growth, this chapter provides a
proposition of the interaction between aid and the recipient country’s
incentive regime and how this interaction may influence economic growth.
Weakness of policy design (i.e., lack of contingency measures to deal with
external shocks, and tensions between policy conditionality and the recipient
country’s national sovereignty), coupled with the problems caused by lack of
state and institutional capability (i.e., lack of enforcement of the rule of law
and policy mismanagement) are the main problems that result in LDCs’
failing to achieve sustainable economic growth. This argument is certified by
performing the simulation of the interactive effects of aid and the recipient
country’s incentive regime on investments, savings, consumption, and
income growth in the aid-growth model. The simulation results demonstrate
that an increase in aid contributes to economic growth when the flow of aid is
stable. In this regard, an increased aid inflow will stimulate the ability to raise
investment and savings along with the path of income growth, although aid is
fungible. In addition, when the flow of aid is unstable or excessive, the model
demonstrates that aid lowers the growth rate of capital accumulation and
hence decelerates economic growth.
Chapter 3__________________________________________________________________ Page 74
From the above point of view, the question as to why policy conditionality
attached to aid might not always promote sustainable economic growth in
Least Developed Countries can be answered as follows:
§
The main cause for aid failure is attributable to the weak design of
policy conditionality and the problems of policy mismanagement
posed by the lack of state and institutional capacity of the recipient
country. Because of these problems the LDCs failed to maintain
economic growth with external viability and this in turn can cause
economic volatility and unstable aid inflows. Therefore, donors’
conditionality to induce changes in policies is not enough to ensure
sustainable economic growth. The aid-growth nexus model has
demonstrated that it is not only policy and institutional changes but
also their design and enforcement that matter for the improvement of
aid effectiveness and the achievement of sustainable economic
growth.
§
As the literature about aid effectiveness predominantly reports a
positive impact of aid on economic growth, it can be argued that the
favourable impact of aid on economic growth is impaired by the
instability of aid flow and economic volatility. These negative effects
are worsened by the weakness of policy conditionality and the lack of
state and institutional capability of LDCs. These are the predominant
factors that cause LDC failure to achieve sustainable economic
growth.
While this chapter has demonstrated that foreign aid could have both
positive and negative effects on economic growth, these results will be
further explored in the case of the Lao PDR. Before carrying out this task,
Chapter 4 addresses the methodological shortcomings of aid fungibility and
aid-growth models, and presents the macroeconometric methods and data
that will be employed to measure the positive effect of stable aid inflows on
economic growth in Chapters 5 and 6. The negative effects of unstable aid
flow on economic growth will be explored in Chapter 7.
Chapter 3__________________________________________________________________ Page 75
Appendix 3.1: Solving the model
Before solving the model, the resource constraint equation is re-written as follows:
s( t ) = y( t ) − c( t ) = ( 1 − π )Φk( t ) − c( t )
Substituting
and
f ( t ) = F0 e µt into
equation (3.1) yields
k&( t ) = ( 1 − π )Φk ( t ) − c( t ) − ( δ + n )k( t ) + F0 e µt
Now the problem of a benevolent social planner can be solved as follows:
Max ∫
c( t )
∞
0
[c( t ) − c ]1−θ − 1 e −( ρ −n )t dt
1 −θ
with respect to the resource constraint:
k&( t ) = ( 1 − π )Φk ( t ) − c( t ) − ( δ + n )k( t ) + F0 e µt
Setting the current value of Hamiltonian gives:
1−θ
[
c( t ) − c ] − 1
H=
+ λ [( 1 − π )Φk ( t ) − c( t ) − ( δ + n )k ( t ) + F e µt ]
1 −θ
0
Deriving the first order conditions gives:
∂H
−θ
= 0 ⇒ [c( t ) − c ] = λ( t )
∂c( t )
(A3.1)
⇒ −θ [c( t ) − c ]
(A3.2)
−θ − 1
c&( t ) = λ&( t )
⇒ λ&( t ) + [(1 − π )Φ − δ − ρ ]λ( t ) = 0
(A3.3)
Use resource constraint k&( t ) and substitute (A3.1) and (A3.3) into (A3.2) to find the
system differential equations:
⇒ c&( t ) = θ −1 [c( 0 ) − c ][( 1 − π )Φ − δ − ρ ]
(A3.4)
⇒ k&( t ) = [(1 − π )Φ − δ − n]k ( t ) − c( t ) + F0 e µt
(A3.5)
Use (A3.3) to find λ( t )
∫e
[(1−π )Φ −δ − ρ ]t
[λ&( t ) + [(1 − π )Φ − δ − ρ ]λ( t )] = 0
Chapter 3__________________________________________________________________ Page 76
⇒ λ( t ) = λ0 e −[(1−π )Φ −δ − ρ ]t
(A3.6)
Substitute (A3.6) into (A3.1) to find the optimal path of consumption c( t )
[c( t ) − c ]−θ
= λ0 e − [(1−π )Φ −δ − ρ ]t
⇒ c( t ) = c + [c( o ) − c ]eθ
[(1−π )Φ −δ − ρ ]t
−1
= C 0 eθ
−1
[(1−π )Φ −δ − ρ ]t
(A3.7)
Substitute (A3.7) into (A3.5) to find the optimal path of capital k ( t )
−1
k&( t ) − [(1 − π )Φ − δ − n ]k ( t ) = −C 0 eθ [(1−π )Φ −δ − ρ ]t + F0 e µt
∫e
− [(1−π )Φ −δ − n ]t
[k&( t ) − [(1 − π )Φ − δ − n]k( t )]dt =
− C 0 ∫ eθ
−1
[(1−π )Φ −δ − ρ ]t −[(1−π )Φ −δ − n ]t
dt + F0 ∫ e µt −[(1−π )Φ −δ −n ]t
− C 0 eθ [(1−π )Φ −δ − ρ ]t −[(1−π ) A −δ −n ]t
F0 e µt −[(1−π )Φ −δ −n ]t
k ( t ) = −1
+
θ [(1 − π )Φ − δ − ρ ] − [(1 − π )Φ − δ − n ] µ − [(1 − π )Φ − δ − n ]
−1
e
− [(1− d ) A −δ − n ]t
− C0 eθ [(1−π )Φ −δ − ρ ]t
F0 e µt
k ( t ) = k + −1
+
θ [(1 − π )Φ − δ − ρ ] − [(1 − π )Φ − δ − n ] µ − [(1 − π )Φ − δ − n ]
−1
][
[
⇒ k ( t ) = k + k( 0 ) − k eθ
where k ( 0 ) − k =
and F1 =
−1
[(1−π )Φ − δ − ρ ]t
+ F1e µt
]
(A3.8)
− C0
θ [(1 − π )Φ − δ − ρ ] − [(1 − π )Φ − δ − n ]
−1
θ −1 [(1 − π )Φ − δ − ρ ] − [(1 − π )Φ − δ − n ]
F0
− C0
µ − [(1 − π )Φ − δ − n ]
where k is the level of subsistent capital and its value is similar to the one set by
Steger (2000), in that k ≡
c
( 1 − π )Φ − δ − n
Use (A3.7) to find the growth rate of consumption at the balanced-growth
equilibrium position: lim
t →∞
c&( t )
c( t )
Differentiating (A3.7) with respect to time, then dividing the result by c(t) gives:
c&( t ) = [c( o ) − c ]θ −1 [(1 − π )Φ − δ − ρ ]eθ
−1
[(1−π )Φ −δ − ρ ]t
c&( t ) [c( o ) − c ]θ −1 [(1 − π )Φ − δ − ρ ]eθ [(1−π )Φ −δ − ρ ]t
=
−1
c( t )
c + [c( o ) − c ]eθ [(1−π )Φ −δ − ρ ]t
−1
Chapter 3__________________________________________________________________ Page 77
Applying the l’ hospital rule gives:
[c&( t )]' → θ −1 [(1 − π )Φ − δ − ρ ]
c&( t )
= lim
t → ∞ c( t )
t → ∞ [c( t )]'
(A3.9)
lim
Use (A3.8) to find the growth rate of capital at the balanced-growth equilibrium
k&( t )
t →∞ k ( t )
position: lim
Differentiating (A3.8) with respect to time, then dividing the result by k(t) gives:
[
][
[
][
−1
k&( t ) = k ( 0 ) − k θ −1 [(1 − π )Φ − δ − ρ ]eθ [(1−π )Φ −δ − ρ ]t + µF1e µt
]
−1
k&( t ) k ( 0 ) − k θ −1 [(1 − π )Φ − δ − ρ ]eθ [(1−π )Φ −δ − ρ ]t + µF1e µt
=
−1
k( t )
k + k ( 0 ) − k eθ [(1−π )Φ −δ − ρ ]t + F1e µt
[
][
]
]
In the balanced-growth equilibrium position the economy would no longer need to
borrow foreign capital, as the economy can rely on domestic savings to finance
investment demands. So, the term f ( t ) = µF1 e µt = 0
Then
][
[
−1
k&( t ) k ( 0 ) − k θ −1 [(1 − π )Φ − δ − ρ ]eθ [(1−π )Φ −δ − ρ ]t
=
−1
k( t )
k + k( 0 ) − k eθ [(1−π )Φ −δ − ρ ]t
[
][
]
]
Applying the l’ hospital rule gives:
[ ]
'
k&( t )
k&( t )
lim
=
→ θ −1 [(1 − π )Φ − δ − ρ ]
t →∞ k ( t )
t → ∞ [k ( t )]'
lim
(A3.10)
Use the system differential equation (A3.4) and (A3.5) and apply the timeelimination method to find the policy function
c&( t ) = [c( t ) − c ]θ −1 [(1 − π )Φ − δ − ρ ]
k&( t ) = [(1 − π )Φ − δ − n ]k ( t ) − c( t ) + F0 e µt
c' ( k ) =
[c( k ) − c ]θ −1 [(1 − π )Φ − δ − ρ ] = c' ( k )θ −1 [(1 − π )Φ − δ − ρ ]
c&
=
[(1 − π )Φ − δ − n] − c' ( k )
k& [(1 − π )Φ − δ − n ]k − c( k ) + F0 e µt
Simplifying the result gives:
c' ( k )[[(1 − π )Φ − δ − n ] − c' ( k )] = c' ( k )θ −1 [(1 − π )Φ − ρ − δ ]
[c' ( k )] = c( t ) − c = [(1 − π )Φ − δ − n] − θ −1 [(1 − π )Φ − δ − ρ ]
k( t ) − k
[
]
⇒ c( t ) = z k ( t ) − k + c
(A3.11)
Chapter 3__________________________________________________________________ Page 78

Find the transitional path of saving rate: s y ( t ) =  1 −

c( t ) 

y( t ) 
Simplifying (A3.11) gives:
[
]
c( t ) = {[(1 − π )Φ − δ − n ] − θ −1 [(1 − π )Φ − δ − ρ ]} k ( t ) − k + [( 1 − π )Φ − δ − n ]k
[
]
[
]
c( t ) = [(1 − π )Φ − δ − n ] k( t ) − k − θ −1 [(1 − π )Φ − δ − ρ ] k ( t ) − k + [( 1 − π )Φ − δ − n]k
− c( t ) = θ
−1
[(1 − π )Φ − δ − n][k ( t ) − k ] − [( 1 − π )Φ − δ − n]k ( t )
In an equilibrium position of the balance of payments the following identity holds:
i( t ) + f ( t ) = s( t ) + f ( t )
[( 1 − π )Φ − δ − n]k ( t ) − c( t ) + f ( t ) = s( t ) + f ( t )
[( 1 − π )Φ − δ − n]k( t ) − c( t ) = y( t ) − c( t )
Deleting c(t) from the above equation then multiplying both sides by
c( t )
yields:
y( t )
[( 1 − π )Φ − δ − n]k ( t ) c( t ) = c( t )
y( t )
Substituting −
c( t )
= s y − 1 into the above equation gives:
y( t )
γ
0
[( 1 − π )Φ − δ − n]k ( t )s y ( t ) = θ −1 [(1 − π )Φ − δ − n][k ( t ) − k ]
sy( t ) =
[(1 − π )Φ − δ − ρ ] k ( t ) − k
θ [(1 − π )Φ − δ − n ] k ( t )
(A3.12)
Find the transitional dynamic of growth rate of per capita capital and output
y&( t ) k&( t )
c( t ) f ( t )
=
= [(1 − π )Φ − δ − n ] −
+
y( t ) k ( t )
k( t ) k ( t )
(A3.13)
Chapter 4__________________________________________________________________ Page 79
Chapter 4
MACROECONOMETRIC METHODS AND DATA: FOREIGN AID
AND ECONOMIC GROWTH
...The inadequate theoretical foundation of models of aid’s
impact on saving, in particular their failure to consider
economy-wide effects, make them a poor basis for conclusions
about aid’s impact...
White (1992b, p. 121)
…The transition from theory to empirical work in
macroeconomics is not always straightforward. The quality of
the data are never as good as one might like, so compromises
have to be made in moving from theory to empirical
specifications…
Fair (1994, p. 39)
4.1 Introduction
Apart from theoretical shortcomings, the persistence of inconclusive
evidence regarding the interaction between aid and policy conditionality and
its
effect
on
sustainable
economic
growth
was
attributed
to
the
methodological weakness of the aid-growth relationship. Chapters 2 and 3
have indicated that aid has a potentially large effect on various
macroeconomic variables such as consumption, savings, investments,
imports, exports, prices and real effective exchange rates. This economywide effect of aid causes some drawbacks to the econometric methods
employed to estimate the aid-growth and aid fungibility coefficients.
McGillivray and Morrissey (2001) indicate that aid fungibility models are
subject to a number of limitations. Overall, aid fungibility models have both
theoretical and methodological problems (ibid.). Hansen and Tarp (2000)
suggest that the estimation of the aid-growth coefficient obtained from the
aid-growth regression has been highly sensitive to the set of control
variables. The study also stresses that more research on theoretical work is
needed before the aid-growth regression can be used for policy purposes.
Chapter 4__________________________________________________________________ Page 80
To tackle the econometric shortcomings of the aid fungibility and aid-growth
regression models, many studies have employed multi-equation models to
capture the macroeconomic impact of aid. Gupta and Lensink (1995)
construct a macroeconometric model to measure the effect of aid fungibility
on economic growth. White (1998) indicates that many studies of aid
effectiveness
employed
computable
general
equilibrium
(CGE)
and
macroeconometric models to analyse the potential effects of aid on
economic growth. Nevertheless, modelling the macroeconomic impact of aid
may not be limited to these two approaches. “There are currently five major
approaches
to
macroeconomic
modelling:
the
traditional
Cowles
Commission structural equations approach, unrestricted and Bayesian
VARs, “structural” VARs, linear rational expectation models, and the
calibration approach associated with real business cycle theories” (Pesaran
and Wickens, 1995, p.1). Selecting one of these approaches requires the
model builder to balance between the role of economic theory and statistics.
In other words, the model builder should decide whether “to estimate a
model derived from formal economic theory, or.. to find a model that accords
well with data?” (ibid., p.1).
Due to the limitation of longer time-series data and the lack of a social
accounting matrix for the Lao PDR, macroeconometric models are employed.
The traditional Cowles Commission structural equations approach is also
adopted for the macroeconometric modelling of aid fungibility and the aidgrowth nexus in this study. The rest of this chapter is organised as follows:
Section 4.2 discusses the limitations of the existing aid fungibility model and
presents the specification of the macroeconometric model of aid fungibility.
Section 4.3 discusses the shortcomings of the aid-growth regression and
presents the macroeconometric model of the aid-growth nexus. Section 4.4
introduces the estimation methods for the macroeconometric model while
Section 4.5 presents the sources of data and methods of variable
construction. Section 4.6 summarises the discussion in this chapter.
Chapter 4__________________________________________________________________ Page 81
4.2 The aid fungibility model
Aid is said to be fungible at the aggregate level if an increase in aid intended
to finance investment does not raise investment to the extent of the value of
aid inflow but enables the recipient government to raise government
consumption spending and/or lower taxes. The existing empirical studies
have mainly employed two distinct methodological approaches to estimate
the coefficient of aid fungibility. However, these approaches have both
theoretical and methodological shortcomings, which make the estimation of
aid fungibility coefficients deficient (Lensink and White 2000, pp. 12-13).
This issue is discussed in Section 4.2.1. Section 4.2.2 presents a demanddetermined macroeconometric model which has been developed to measure
the impact of aid fungibility on domestic investment. The model is also
developed to tackle the defects of the existing approaches to aid fungibility
issues.
4.2.1 Limitations of the existing aid fungibility model
As discussed in the “fiscal response” literature, there are two main
approaches to the estimation of the aid fungibility coefficient. Although the
fundamental deficiency of aid fungibility is related to the way in which
theoretical models have been established, the limitation of aid fungibility
model discussed here focuses on the weakness of empirical models, as this
is relevant to the macroeconometric model of aid fungibility developed in the
next section.1
To discuss the shortcomings of these approaches, the
empirical framework of aid fungibility discussed in Chapter 2 is reproduced
in this section. The first approach is represented by two sets of system
equations. The first set captures the recipient government’s fiscal behaviour
in response to aid inflow. This is illustrated by equations (4.1) through to
(4.4) as follows:
1
For more detailed discussion of the limitations in both theoretical and empirical model of fungibility, see McGillivray and
Morrissey (2001, Sections II and III)
Chapter 4__________________________________________________________________ Page 82
DI t = f1 ( DI t* , Tt , At )
(4.1)
IDI t = f 2 (CGt* , IDI t* , Tt , At )
(4.2)
CGt = f 3 (CGt* , IDI t* , Tt , At )
(4.3)
Tt = f 4 (Tt * , CGt* , CGt , DI t , At )
(4.4)
where the fiscal-choice variables are direct public investment (DI), indirect
public investment (IDI), government consumption spending (CG) and
government revenue (T). The variables that influence the fiscal-choice
variables are total aid flow (A), which can be classified into grant aid and
loan aid. An asterisk (*) denotes the targeted fiscal variables. These
variables form the second set, and are defined in equations (4.5) through to
(4.8) as follows:
DI t* = f 5 (Yt −1 , IPt )
(4.5)
IDI t* = f 6 ( Enrt , Yt , GYt )
(4.6)
CGt* = f 7 (CG )
(4.7)
Tt* = f 8 ( Yt , M t −1 )
(4.8)
where Y is real income and GY is growth of real income, IP is private
investment, Enr is primary school enrolments and M is imports.
There are two main problems with this approach. The first problem is related
to the estimation of the aid fungibility coefficient. This problem will occur if
the fiscal-choice variables (Eq. 4.1-4.4) are closely related to the explanatory
variables of the regression of targeted fiscal variables (Eq. 4.5-4.8). In this
context, if the regression of the targeted fiscal variables has R2 close to 1,
and then the estimation of (Eq. 4.1-4.4) will be regressing the fiscal choice
variables on themselves (e.g., if DI ≈ DI*, IDI ≈ IDI*, CG ≈ CG*, T ≈ T*).
Therefore, any aid fungibility coefficients obtained from the estimation of
fiscal choice regressions will provide inconclusive information of the
Chapter 4__________________________________________________________________ Page 83
government’s fiscal behaviour in response to aid inflows (White, 1994, pp.
159-160).
The second problem is the lack of dynamic behaviour in the structural
equations so the model does not allow for the full effect of the aid on fiscal
behavioural variables in the short run and long run. “Even if the feedback
effect through the target variables were incorporated, analysis based on this
model would remain a partial one since it contains no economic feedback
mechanisms” (ibid., 1994, p. 162). Franco-Rodriquez et al. (1998) and
Franco-Rodriquez (2000) attempt to solve these problems by applying the
cointergrating technique to obtain the values of the target variables and
using a reduced form of the choice-variable equation system to obtain shortrun and long-run fiscal behaviour. However, the model examines the aid
fungibility issue by focussing only on the government sector thereby
neglecting the economy-wide effects of aid (McGillivray and Morrissey,
2001, p. 30).
The empirical framework of aid fungibility for the second approach is
represented by the equation set (4.9) through to (4.12) as follows:
Tt = f 9 ( At , Z t )
(4.9)
CGt = f10 ( At , Z t )
(4.10)
IGi ,t = f 11 ( Ai ,t , A≠i ,t , Z t )
(4.11)
n
CGt + ∑ IGi ,t = Tt + At + PSBRt
(4.12)
i =1
where T, and CG are defined as above, IG is public investment, PSBR is the
public sector borrowing requirement, Ai,t is aid allocation to sector i at time t,
, A≠i,t is aid allocation to other sectors at time t and Z is a vector for control
variables. Pack and Pack (1990; 1993) use GDP as the control variable,
while Feyzioglu et al. (1998) use the infant mortality rate, average years of
Chapter 4__________________________________________________________________ Page 84
schooling in the labour force, average ratio of a neighbouring country’s
military expenditure to GDP, and ratio of agricultural output to GDP.
Equation (4.11) captures the sectoral aid fungibility coefficients so that aid
fungibility at the aggregate level can be obtained by summing up the sectoral
aid fungibility coefficients. This approach shares a common problem with the
first approach, in which the lack of dynamic structure in the model makes it
only a partial analysis of the impact of aid inflow on fiscal behaviour. Also,
this approach analyses the aid fungibility issue by focusing only on the
government sector.
4.2.2 A macroeconometric model of the aid fungibility
To tackle the abovementioned shortcomings of the aid fungibility model, it is
necessary to incorporate the economy-wide effects of aid into the
macroeconomic framework. In this regard, the interplay between various
economic sectors and economic feedback mechanisms can be used to track
the potential impact of aid inflows on various macroeconomic variables.
Although aid may affect both the demand side and the supply side of the
economy, the modelling of the impact of aid fungibility on investment is
focused on the demand side of the economy.2 This is because aid may have
some influence on the decisions relating to demand for consumption,
savings and investment in both public and private sectors through the impact
of aid inflows on economic prices and income. In this context, a
macroeconometric model of aid fungibility for the case of the Lao PDR is
developed from the structure of the Keynesian approach to macroeconomic
modelling. The model explicitly allows for the interaction between the public
and private sectors. Although the external sector is included in the model, it
is treated as an exogenous variable and is allowed to affect both the public
2
Gupta and Lensink (1995) construct a macroeconometric model of fungibility with heavy emphasis on the impact of aid
on the demand side of the economy. The model explicitly allows for interactions between four sectors in response to aid
inflows. Those sectors are the non-bank sector, the government sector, the banking sector and the external sector. If in
the model, aid is only given to only the government, the simulations show that partial fungibility has a positive effect on
public investment and very small effect on GDP In the same model, the simulations also show that full fungibility has a
negative effect on public investment but still has a very small effect on GDP (Gupta and Lensink, 1995, p.66).
Chapter 4__________________________________________________________________ Page 85
and private sectors through the transmission mechanism in the equilibrium
condition of the model.
The macroeconometric model of aid fungibility for the case of the Lao PDR
consists of eight equations, in which five are stochastic (i.e., equation 4.134.17) and three are definitional and equilibrium conditions (i.e., equation
4.18-4.20). The specifications of the model are expressed in a generic form
as follows:
Macroeconometric model of aid fungibility
Behavioural equations
Public sector
CGt = f 1 (Tt , At , INFt )
IGt = f 2 (Tt , At , INFt )
(4.13)
(4.14)
Tt = f 3 (Yt , At , FDI t , INFt )
(4.15)
Private sector
IPt = f 4 ( IG, YDt , FDI t , INFt )
(4.16)
(4.17)
CPt = f 5 ( Yd ,t , INFt )
Identity equations
=
*
t
YDt = Yt + NTRt
(4.18)
(4.19)
ADt = ASt = Yt
(4.20)
DI
t
f 1 ( DI
,Tt , A
t
)
where CG: government consumption spending,
IG: government capital expenditure,
T : government revenue,
IP: private investment,
CP: private consumption,
YD: disposable income,
Y: income,
NTR net transfer payments,
A: total aid inflows,
FDI: foreign direct investment,
INF: inflation,
NX: trade balance,
AD: aggregate demand, and
AS: aggregate supply,
Chapter 4__________________________________________________________________ Page 86
In the public sector, aid inflow (A) is included in the functions of the fiscal
choice variables (i.e., equation (4.13) to (4.15)), to capture the influence of
aid on the government’s fiscal behaviour. This is in line with the model
employed by Pack and Pack (1990, 1993). Apart from aid inflow, the Lao
Government has relied on government revenue (T) to finance public
consumption (CG) and provided the counterpart funds for public investment
(IG). In this regard, T is included as an explanatory variable in equation
(4.13) and (4.14). Furthermore, the Lao Government has mainly derived tax
revenue from income tax, excise tax and other non-income tax (i.e., profits
tax, turnover tax, trade tax, and royalties). Foreign firms have been a main
source for non-income tax. Thus, in equation (4.15), foreign direct
investment (FDI) is used as a proxy variable for non-income tax, while
income (Y) is used as proxy variable for income tax and excise tax. Inflation
(INF) is included in equations (4.13) to (4.15) to capture the feed back effect
due to inflation tax policy.3
In the private sector, investment function (equation (4.16)) is derived from
the accelerator theory, which “asserts that investment spending is
proportional to the change in output” (Dornbusch and Fischer, 1994, p.348).
Thus, disposable income (YD) is included as a proxy variable for the change
in output. It is assumed that the investment function is dominated by fiscal
policy, foreign private capital inflow and macroeconomic instability. As
discussed in the fiscal response literature in Section 2.3.2, public investment
(IG) is therefore included to capture the crowding out (or crowding in)
effects. The inflow of foreign direct investment (FDI) is also used as an
explanatory variable because private investment in the Lao PDR is
dominated by the FDI inflows. Macroeconomic instability may create an
uncertain environment for investment decisions, which in turn causes the
3
Inflation tax is an implicit tax levied by the government by means of base money creation (e.g., printing money). Inflation
tax has long been recognised as an important source of government revenue in many developing countries. This is no
exception for the Lao PDR. Inflation tax can have a positive and negative effect on real government revenue. In this
context, inflation is used as an explanatory variable linking the inflation tax and the collection of tax revenue. For further
detailed discussions on inflation tax see Agenor and Montiel (1996, pp.111-121).
Chapter 4__________________________________________________________________ Page 87
delay or even the cancellation of investment decisions (Dixit and Pindyck,
1994). To capture the effect of macroeconomic instability on private
investment, inflation is used as a proxy variable for macroeconomic
instability. With respect to consumption, the consumption function (CP) is
derived from the Keynesian consumption function, which asserts that
consumption increases in proportion to disposable income (YD). The inflation
variable is also used as an explanatory variable because it can have a direct
effect on real consumption, as the consumers suffer from money illusion or if
money enters a consumer’s utility function (see for example Deaton, 1977;
Juster and Wachtel 1972a, 1972b; and von Furstenburg, 1980).
Note that the real interest rate could be included in the consumption and
investment functions. However, owing to less developed financial and
banking system in the Lao PDR, residents would have less opportunity to
smooth their consumption path. Moreover, the prevailing informal loan
market and the controlling of interest rates by the Lao Government make it
hard to observe the impact of the official interest rate on the demand for
investment in the private sector. Therefore, the potential effects of real
interest rate on consumption and investment have been excluded.
With respect to identity equations, equations (4.18) and (4.19) represent the
national account identity for aggregate demand and disposable income,
respectively. The final equation (4.20) equates aggregate demand,
aggregate supply and income.
4.3 The aid-growth nexus model
Apart from the theoretical critiques of the aid-growth nexus pointed out in the
two-gap model, modified aid-growth regressions to tackle the critiques
discussed in Chapter 2 have also been problematic because many empirical
studies have failed to provide statistically significant aid-growth coefficients.
This perceived lack of evidence on the aid-growth nexus has generally been
blurred by associated critiques of the econometric methodology applied in
Chapter 4__________________________________________________________________ Page 88
such studies. This issue is discussed in Section 4.3.1. Section 4.3.2
presents the macroeconometric model of the aid-growth nexus for the case
of the Lao PDR.
4.3.1 Econometric issues of the aid-growth regression
As discussed in Chapter 2, many of the aid effectiveness studies have
derived their empirical frameworks from various growth theories. In many
cases the reduced form of the aid-growth equation has been employed to
assess the effectiveness of aid. Although many researchers have been
aware of the weakness of this approach and have employed more
sophisticated techniques to tackle the potential econometric misspecification
of the aid-growth regression, the resulting estimation of aid-growth
coefficients have still been ambiguous. This controversial result has been
seen as a problem arising from the econometric misspecification of the aidgrowth model that may have caused the estimated aid-growth coefficient to
be biased and unstable. White (1998) claims that the aid-growth regression
has been misspecified on three counts: (i) omitted variable biases, (ii) single
equation estimations of simultaneous relationships, and (iii) parameter
instability (White, 1998, pp. 27-29). The explanations of why these
specification errors are likely to be present in the aid-growth regression are
presented below. Some of the aid-growth regressions presented in Chapter
2 have been reproduced here for the sake of convenient discussion.
S
Y&
A
OF
= α0 +α1 +α2 d +α 3
+ε
Y
Y
Y
Y
(4.21)
Y&
A
S
OF
= α 0 + α1 + α 2 d + α3
+ α 4Z + ε
Y
Y
Y
Y
(4.22)
where
Y&
A S
OF
is the rate of output growth, , d ,
are aid inflows, domestic
Y
Y Y
Y
savings and other sources of capital inflow as a percentage of GDP,
respectively. Z is a vector for control variables including the growth rate of
various factor inputs and policy variables affecting growth.
Chapter 4__________________________________________________________________ Page 89
As indicated in Chapter 2, Section 2.2, equation (4.21) here may be said to
be misspecified, and may suffer from omitted variable bias. Levine and
Renelt (1991) report that there are over 50 variables found to be significantly
correlated with growth. As such the first type of specification error can be
due to the correlated variables that have been excluded in the regression
equation, clearly the case in equation (4.21).
In an attempt to solve the above specification error more explanatory
variables have been added. Many researchers have done this by employing
equation (4.22) to isolate the effect of aid on growth from the added
explanatory variables.4 However, equation (4.22) is likely to suffer from the
second case of specification error if the variables included are correlated
with both aid and growth. For example, a case in the literacy variable
included by Mosley et al. (1987) indicates higher literacy, which leads to an
increase in growth due to the high quality of labour force. This may result
from an increase in government expenditure on education, which in turn is
also financed by aid. This implies that aid inflows, literacy and growth are
correlated. White (1998) has also pointed out some variables that correlate
with aid and growth; i.e. military expenditure, educational indicators and
macro-policy variables. Moreover, after 1980, the economic policies of
developing countries have been improved as a result of aid expenditure on
policy advice. Easterly (2000) demonstrates that proxies for policy variables
such as financial depth and real overvaluation as well as policy-related
factors like health, education, fertility and infrastructure are highly significant
in the growth regression.
It is obvious that attempts to isolate the effect of aid on growth from the
added explanatory variables would create simultaneous relationships among
4
In the aid-growth model the investment variable is replaced by the ratio of aid to GDP. In this context, the aid variable is
perceived as a factor contributing to economic growth via the accumulation of physical capital and the productivity of
capital. This is equivalent to the “level effect” and “growth effect” in the theory of economic growth presented by Lucas
(1988). Other explanatory variables are employed to capture other factors affecting economic growth via the “level of
efficiency” or the “growth effect” (see Sala-i-Martin, 1997a&b for a similar interpretation of the role for other explanatory
variables in the growth model).
Chapter 4__________________________________________________________________ Page 90
aid, economic growth and the added explanatory variables. These
simultaneous relationships cannot be captured in the aid-growth regression
equation. Thus, including more explanatory variables in the aid-growth
regression analysis (such as the literacy rate in the Mosley et al. (1987)
study of policy variables influenced by aid) is a specification error case of
single equation estimation of simultaneous relationships.5
Lastly, the most serious case of econometric misspecification arises due to
the diversity of aid given to developing countries, in which the impact of aid
on growth can vary with the types of aid and with its spending on diverse
sectors.6 This may cause the estimate of the aid coefficient in the aid-growth
regression to be unstable. White clarifies this point by asserting that:
the impact of aid will be different if it is used for building primary
health clinics rather than for roads, extension services or an
industrial line of credit. All these things should be expected to
increase growth, but through different channels and certainly
with different lag structures.
White (1992b, p. 129)
In addition, White (1998) claimed that “there is no theoretical foundation
whatsoever for the assumption that the impact of aid on growth is constant
either across countries or across time” (ibid., p. 29). It is obvious that the
diversity of the impact of aid on growth mentioned above may also vary
across countries and time. Even if the lag structure effect of aid is
incorporated in the aid-growth regression, it cannot tackle the diversity of the
impact of aid on growth “since the required lag structures will change as the
sectoral composition of aid changes” (ibid, p. 29). Therefore, the estimation
of the aid-growth regression with a sample of cross-country or panel data
may yield unstable aid-growth coefficient.
5
The simultaneity issue may arise if aid is endogenous as a result of the process of aid allocation. However, donors may
conceivably allocate aid by income per capita or their interests in the recipient country rather than economic growth or
income. In this sense, aid will not cause the issue of simultaneity in the aid-growth regression (White, 1992b, p.130).
6
Grant aid with tied conditions lowers the transferred value of aid (Cassen, 1994, p.16).
Chapter 4__________________________________________________________________ Page 91
4.3.2 A macroeconometric model of the aid-growth nexus
To measure the influence of the interactive effects that aid and policy
conditionality may have on economic growth, the standard neo-classical
measure of growth has been modified to integrate several aspects of the
impact of capital flow on the economy through its role in financing domestic
investment, as explained by the two-gap model. Unlike the typical approach
using a single regression equation, the macroeconometric model tracks the
dynamic impact of capital flows through the direct and `indirect effects on
investment, international trade and growth. The impact of policy conditions
on economic growth is captured in the short-run and long-run coefficients of
the aid-growth nexus. By doing this, the model has an advantage of
capturing simultaneous relationships among capital flows and other
endogenous variables in the model. It also reduces the potential
multicollinearity errors that can occur if the aid-growth regression equation is
applied.
The macroeconometric model of the aid-growth relationship consists of five
behavioural equations and three identity equations. The model is presented
in a generic form as follows:
A macroeconometric model of the aid-growth nexus
Behavioural equations
Output growth
GYt = f1 ( IYt ; GLEt )
(4.23)
Investment
IGYt = f 2 (GYt ; AYt , INFt )
(4.24)
IPYt = f 3 (GYt , IGYt ; FDIYt , INFt )
International trade
(4.25)
XYt = f 4 ( MKYt ; YFt REERt )
MKYt = f 5 ( IYt ; REERt )
(4.26)
(4.27)
Chapter 4__________________________________________________________________ Page 92
Identity equations
IYt = IGYt + IPYt
(4.28)
MYt = MKYt + MCYt
CAYt = XYt − MYt + U t
(4.29)
(4.30)
where GY: growth rate of income,
IY:
the share of investment in GDP,
IGY: the share of public investment in GDP,
IPY: the share of private investment in GDP,
XY:
the share of exports to GDP,
MKY: the share of imports of capital goods in GDP,
MCY: the share of imports of consumption goods in GDP,
MY: the share of imports to GDP,
FDIY: the ratio of total foreign direct investment to GDP,
CAY: the ratio of current account deficit to GDP,
AY:
the ratio of total aid to GDP,
INF: the inflation rate,
GLE: the growth rate of effective labour force,
YF:
the weighted average by trade share of growth in per
capita income of the trading partners,
REER: real exchange rate,
U:
the error term including net service transfers.
The behaviour of output growth, as indicated by equation (4.23), is based on
the neo-classical growth model modified to include the ratio of total
investment to GDP and the growth rate of effective labour force.7 Total
investment is divided into public investment (equation (4.24)) and private
investment (equation (4.25)). The factors determining investment decisions
employed in the aid-growth models are similar to those used in the aid
fungibility model, except that the growth of output (GY) is used as a proxy
variable for government revenue (T) and the accelerator effect for public and
private investment functions, respectively.
7
The behaviour of output growth can be derived using the production function Yt = Yt ( K t , LEt ) as follows:
Differentiating Yt
∂Y
∂Yt
w.r.t time gives Y&t = t K& t +
LE&t , then dividing both sides of the equation by Yt and
∂K t
∂LEt
substituting K& t = I yields:
Y&t
∂Y t I t
∂Y t LE t LE& t
Y&
I
L E&
⇒ GYt = f1( IYt ,GLEt ) where GYt = t , IYt = t , GLEt = t .
=
+
Yt
Yt
LE t
Yt
∂K t Y t
∂LE t Y t LE t
Chapter 4__________________________________________________________________ Page 93
With respect to international trade, exports and capital imports are used to
capture the potential effects of foreign capital inflows. As mentioned in
Chapter 3, the LDCs’ ability to raise investment with external viability heavily
depends on international trade and foreign capital inflows. This is no
exception for the Lao PDR. Thus, the export function, as indicated by
equation (4.26), is expected to be dominated not only by economic growth in
the Lao PDR trading partners (YF), but also by the ability to import capital
goods (MKY). The latter, as indicated by equation (4.27), is determined by
total investment (IY) which in turn depends on the inflow of both aid and
foreign direct investment as specified in equation (4.24) and (4.25). The real
effective exchange rate is included in equation (4.26) and (4.27), to capture
the “Dutch disease” effect.
As for the identity equations, equation (4.28) and (4.29) represent total
investment and imports, respectively. Equation (4.30) represents the current
account balance, set equal to the sum of trade account deficits and net
service transfers. In the case of the Lao PDR, net service transfers have less
influence on the movement in the balance of payments. For the sake of
expository simplicity, net service transfers are assumed to be zero.
Therefore, the balance of payments becomes the trade balance. This
equation is included in the model to capture any repercussions from
international trade. It is also employed to address the potential effects of
policy conditionality on the Lao PDR’s ability to raise investment with
external viability.
4.4 Estimating, testing and analysing the macroeconometric
models
Once a macroeconometric model has been specified, the reduced form can
be expressed in matrix form as follows.
A0 xt = A1 zt + A2 zt-1 + A3 xt-1+ et
(4.31)
Chapter 4__________________________________________________________________ Page 94
In this expression xt is the (n x 1) vector of endogenous variables, zt is the (m
x 1) vector of exogenous variables, xt-1 and zt-1 are respectively the xt and zt
vector lagged one period, and et is the (n x 1) vector of residuals. A0 is the (n
x n) matrix of the coefficients of the endogenous variable; A1 is the (n x m)
matrix of the coefficients of the exogenous variables. A2 is the (n x m) matrix
of the coefficients of the lagged exogenous variables. In the A2 matrix, the
entry of intercepts into the structural equations is accommodated by
including a variable which always takes the value of unity. A3 is the (n x n)
matrix of the coefficients of the lagged endogenous variables. It should be
noted that equation (4.31) is dynamic, simultaneous and has error terms that
may be correlated across equations and with their lagged values.
There are many issues associated with the process of selecting methods to
estimate, test and analyse the macroeconometric model. First, the model
builder must select methods that can tackle the econometric issues in time
series and multiple equations simulation. Second, when individual equations
are combined as a system of simultaneous equations, specific criteria are
required for testing the goodness of fit and properties of the model. Finally,
there are various ways the macroeconometric model can be analysed.
These issues are discussed in the following sections.
4.4.1 Estimation methods
Theoretically, estimation methods for macroeconometric models can be
divided into two groups. The first group employs single equation methods
that do not take into account the correlation of error terms across equations.
These estimation techniques consist of Ordinary Least Squares (OLS),
Indirect Least Squares (ILS), Two-Stage Least Squares (2SLS) and Limited
Information Maximum Likelihood (LIML). The second group employs system
equation methods that do account for the error terms across equations.
These estimation techniques consist of Seemingly Unrelated Regression
Estimation (SURE), Three-Stage Least Squares (3SLS), and Full Information
Chapter 4__________________________________________________________________ Page 95
Maximum Likelihood (FIML).8 When applying the estimation techniques from
both groups to time series data, the model builders have to make sufficient
stationary assumptions about the variables employed in the model. However,
the majority of macroeconometric model builders either explicitly or implicitly
assumed variables are trend stationary (cointergrated). A time trend is
normally included to test whether the estimation is picking up spurious
correlation from common trending variables. If including a time trend in the
equation substantially changes some of the estimated coefficients, this is
cause for concern.9
In practice, if the sample size is small, utilising system methods for the
estimation could have more problems than using single-equation methods.
The problems could arise if either the number of predetermined variables
exceeds the number of sample points or the number of behavioural
equations exceeds the number of sample points (Challen and Hagger, 1983,
p. 128). To tackle these problems, single equation methods are frequently
used. The ILS technique is appropriate, if the structure of a model is just
identified while the 2SLS technique is appropriate if the structure of the
model is overidentified (Gujarati, 1995, pp. 678-700). The OLS technique is
appropriate, if either the structure of the model is recursive or the variables
in a behavioural equation turn out to be cointergrated. In the latter case,
utilising the OLS technique yields a super consistent estimator for the
cointergrating parameters of those variables (Enders, 1995, p. 374).
However, there is a danger of small-sample biases in OLS estimates
(Banerjee et al., 1993, pp. 214-230).10
From the above discussion, the estimation problem that is likely to be
encountered in this study is the small sample size to estimate the
8
For more detailed discussions on estimation methods see for example Challen and Hagger (1983), Pindyck and
Rubinfeld (1991).
9
The estimation issues with non-stationary data are discussed in Fair (1994).
10
Although 30 observations are recommended, the suitable sample size also depends on the characteristics of the
problem at hand for estimators to display their asymptotic properties. Goldfeld and Quandt (1972, p. 277) report an
example in which a sample size of 30 is sufficiently large and another example in which a sample of 200 is required.
Chapter 4__________________________________________________________________ Page 96
cointergrating coefficients of the behavioural equations. This is because only
20 annual observations are employed for the estimation. Also, variables
included in the behavioural equations are likely to be cointergrated among
themselves through a direct and/or indirect link from either the identity
equations or the specification of behavioural equations of those variables.
Therefore, the estimation strategy adopted here is an ad hoc technique to
handle the abovementioned problems. The autoregressive Distributed Lag
(ARDL) approach to cointegration developed by Pesaran and Shin (1995) is
used to estimate the dynamic structure of behavioural equations.
The ARDL procedure involves two stages. In the first stage, each
behavioural equation is transferred to the error correction form of the
underlying ARDL model. The error correction version of the ARDL (p, q) in the
variables Xt and Zt is given by:
p
q
i=1
j=1
∆X t = α + ∑ Ai ∆ X t -i + ∑ Bi ∆Z t - j + C 0 X t -1 + C1 Z t -1 + u t
(4.32a)
where Xt is an endogenous variable, α is an intercept, Zt is an explanatory
variable and µt is the random error term.
The F-statistic for the joint test of the coefficients C0 and C1 is computed to
test for the long-run relationship between Xt and Zt. The null hypothesis is
that the coefficients C0 and C1 in equation (4.32a) are jointly equal to zero. In
other words, the null hypothesis states that there is no long-run relationship
between Xt and Zt. Then, the computed F-statistic is compared with the
critical value bounds of the F-statistic that Pesaran et al. (1996) have
tabulated. If the computed F-statistic is higher (lower) than the upper (lower)
bound of the critical value of F-statistic, the null hypothesis would be
rejected (accepted).
In second stage, if variables in each behavioural equation are found to be
cointergrated (i.e., the null hypothesis is rejected), the ARDL method can be
employed to estimate the dynamic structure of the behavioural equations
Chapter 4__________________________________________________________________ Page 97
using the OLS method. The estimations proceed “without needing to know
whether the underlying variables are I(0) or I(1)” (Pesaran and Pesaran,
1997, p. 304).11
The dynamic structure of the ARDL (p, q) model takes the following form:
p
q
i=1
j=0
X t = α + ∑ Ai X t -i + ∑ Bi Z t − j + µ t
(4.32b)
where Xt is an endogenous variable, α is an intercept, Zt is a vector of
explanatory variable, p and q are respectively a number of the lag length of
Xt and Zt, and µt is the random error term.
As mentioned earlier, there is a danger of small-sample biases in OLS
estimates. To obtain coefficients that accord well with the models, much
experimentation must be performed. For example, different functional forms
are tried, and right hand side (RHS) variables are dropped if they have
coefficient estimates that are not of the expected sign. Despite the
coefficients being statistically insignificant, some explanatory variables are
retained in each equation, if they improve the tracking ability of the
simulation model.12 It should be noted that the selection for lag length (i.e. i
and j) should take many factors into consideration. For example, the lag
length should be consistent with the theory and data admissible (Hendry and
Richard, 1983, p. 3-33). Since only 20 annual observations are used for the
estimation, the lag length is restricted to one lag. The computer software
employed for the estimation is Microfit. For the simulation, SAS and Eview
are employed.
11
This method avoids the requirements of pre-testing of the order of integration, which is necessary in other
cointegration methodologies. Also, this method avoids the problems of serial correlation that arise in the residual-based
cointegration methods by an appropriate augmentation (Pesaran, et. al, 1996).
12
The main purpose of building the macroeconometric models in this study is to measure the magnitude of the impacts
of aid on economic growth. In this context, the tracking ability of the models is important. The more the model is able to
track the trends of historical data, the better the multiplier analysis is able to capture the Lao PDR’s economic behaviour.
Chapter 4__________________________________________________________________ Page 98
4.4.2 Goodness-of-fit testing methods
Once the estimation of behavioural equations is completed and before the
model can be used for various analyses, it is subjected to a number of tests.
The objective of this step is to determine whether the models reach an
acceptable standard of performance. Essentially, when single equations are
combined as a system of simultaneous equations, the goodness of fit for
each single equation does not always imply that the model will have a good
tacking ability for the dynamic path of the historical data. As Pindyck and
Rubinfeld pointed out:
In a multiple-equation model each individual equation may have
a very good statistical fit, but the model as a whole may do a
poor job in reproducing the historical data. The converse may
also be true; the individual equations of a simulation model may
have a poor statistical fit, but the model as a whole may
reproduce the historical time series very closely.
Pindyck and Rubinfeld (1991, p. 332)
In this regard, the goodness-of-fit for the macroeconometric system can be
evaluated by using an ex-post simulation or historical simulation (i.e. a
comparison of the historical data series with the one that is predicted by the
model). To do this requires solving the macroeconometric systems in
equation (4.31), by either the stochastic simulation method or the determistic
simulation method. The latter method is applied in this study. Thus, the error
term is assumed to be zero (i.e., E(et)=0). The deterministic simulation
solution (the simulated value) is given by:
x ts = Π10 z t + Π 11 z t −1 + Π 21 x t −1
(4.33)
where Π10=A0-1A1, Π11= A0-1A2, Π21= A0-1A3
To indicate the goodness-of-fit for the model and its predictive ability,
several statistics will be calculated. In this regard, Simulation Mean Error
(SME), Simulation Mean Percent Error (SMPE), Root Mean Square Error
Chapter 4__________________________________________________________________ Page 99
(RMSE), Root Mean Square Percentage Error (RMSPE) and Theil’s Inequality
Coefficient (U) are employed. The formula of these statistics are given by:
SME =
1
T
SMPE =
∑ (x
T
)
(4.34)




 x ts − x ta

 xa
t =1 
t
T
∑
1
T
∑ (x
T
1
T
1
T
t =1
− x ta
(4.35)
)
2
(4.36)
 x ts − x ta

 xa
t =1 
t
T
∑
∑ (x
T
s
t
− x ta
t =1
∑( )
T
s
t
t =1
1
RMSPE =
T
1
T
− x ta
t =1
RMSE =
U=
s
t
2
x ts
1
+
T




2
(4.37)
)
2
∑( )
T
(4.38)
2
x ta
t =1
The numerical value of U falls between 0 and 1. If U =0, it means there is a
perfect fit. If U=1, on the other hand, the predictive performance of the model
is as bad as it possibly could be. In addition, the closer the values of RMSE,
RMSPE and U are to zero, the higher will be the tracking ability of the model.
Apart from such statistical evaluation, the plot between the actual data ( x ta )
and the simulated values ( x ts ) will be used to check the tracking ability of the
turning point of the model’s endogenous variables.
4.4.3 Methods for analysing macroeconometric models
A useful way to examine the properties of macroeconometric models is to
consider how much the predicted values of the endogenous variables
change when one or more exogenous variables change. This exercise is
called “multiplier analysis”, and there are two types of multiplier analysis.
First, the “impact” or “short-run” multiplier is used to measure the change in
th
the period t value of the n endogenous variable per unit increase in the
Chapter 4__________________________________________________________________ Page 100
period t value of the mth exogenous variable, with all other determinants of
endogenous variable held constant. Second, the “long run multiplier” is used
to measure the total change in the value of the nth endogenous variable per
unit maintained increase in the period t value of the mth exogenous variable.
To compute the multipliers requires deriving the reduced form of the linear
macroeconometric system and can be done in the following procedure. Rewrite the reduced form of the macroeconometric system obtained from (4.33)
as follows:
x t = Π 10 z t + Π 11 z t −1 + Π 21 x t −1
(4.39)
Re-arrange the system into other forms:
 x t  Π 10
x  =  0
 t −1  
Π 11   z t  Π 21
+
0   z t −1   I
Π
 x 
Setting x *t =  t  ,Π*10 =  10
 x t −1 
 0
0  x t −1 
0  x t − 2 
(4.40)
Π 11  *  z t  *
0 *
 x t −1 
Π
, z t   ,Π 21 =  21
, x t −1 = 


 gives
0   z t −1 
0
 I
 x t −2 
x *t = Π*10 z*t + Π*21 x *t −1
(4.41)
Lagging (4.41) by one period gives:
x *t −1 = Π *10 z *t −1 + Π *21 x *t − 2
(4.42)
Substituting (4.42) into (4.41) gives:
( )
2
x *t = Π*10 z *t + Π *21Π *10 z *t −1 + Π*21 x *t −2
(4.43)
Lagging (4.42) by one period and substituting the result into (4.43) gives:
( )
( )
2
3
x *t = Π *10 z *t + Π *21Π *10 z *t −1 + Π *21 Π *10 z *t − 2 + Π *21 x *t −3
(4.44)
Continuing in this way for t-1 substitutions gives:
( )
2
x *t = Π *10 z *t + Π *21 Π *10 z *t −1 + Π *21 Π *10 z *t − 2
( )
3
( )
+ Π *21 Π *10 z *t −3 + ... + Π *21
t −1
( )
t
Π *10 z1* + Π *21 x *0
th
(4.45)
From equation (4.45) the n,m impact multiplier is defined as the first period
of dynamic multipliers and takes the following value:
Chapter 4__________________________________________________________________ Page 101
∂x tn
∂z tm
= Π *10nm
(4.46)
The long run multiplier is defined as the total sum of dynamic multipliers,
which will exist if and only if all of the characteristic roots of Π*21 are less
than unity in absolute value. Therefore, the n,mth long run multiplier takes
the following value:
∂x tn
∂z tm
∂xtn
∂z tm
Other
( )
( )
Π * + Π * Π * + Π * 2 Π *
10
21 10
21
10
= lim 
t →∞ 
* 3
*
+ Π 21 Π 10 + ... + Π *21

(
=  I − Π *21

)
interesting
−1


t −1 *
Π 10 
(4.47)
( )
Π *10 

and
(4.48)
productive
methods
of
analysing
the
macroeconometric system are counterfactual analysis and Goal-seeking
analysis. A counterfactual analysis deals with questions of the “what if”
variety (experiments) and solving for right hand side variables (goalseeking). The “what if” analysis shows how alternative assumptions of the
change in exogenous variables could affect a change in endogenous
variables. For example, if foreign aid increases by 10 percent from its
historical trend what will happen to prices and output? Goal-seeking analysis
shows how much the exogenous variables have to change in order to
achieve some target change in endogenous variables. For example, what
changes in aid inflow could boost economic growth to 7 percent per annum?
It should be noted that the abovementioned techniques are subject to the
“Lucas critique” because the coefficients of the specified equation that
capture the aggregate relationships could be expected to vary with changes
in policy rules and the structure of the economy. Therefore, these techniques
could invalidate the policy conclusions derived from the model. As Lucas
explains:
…given that the structure of an econometric model consists of
optimal decision rules of economic agents, and that optimal
decision rules vary systematically with changes in the structure
Chapter 4__________________________________________________________________ Page 102
of series relevant to the decision maker, it follows that any
change in policy will systematically alter the structure of
econometric models.
Lucas (1976, p. 41)
In avoiding this problem, Lucas proposed including policy coefficients and
disturbance variables into a model so that any policy change can be
analysed by altering these two variables. While the Lucas policy evaluation
critique is a plausible principle, “an extensive search of the literature reveals
virtually no evidence demonstrating the empirical applicability of the Lucas
critique” (Ericsson and Irons, 1995, p. 301). Sims (1982, 1986) actually
ignores the “Lucas critique” by arguing that
…agents with rational expectations understand the choices
facing policymakers and so form probability distributions over
the range of possible policy stances. Thus, what happen to be
regime changes are, in fact, particular draws from probability
distributions that have already been integrated into the
optimisation problems of individual agent. A true regime change
might be the change of the entire distribution. But not only are
such changes rare, the range of possible distribution might
itself be governed by an even higher-order distribution.
Sims (1982, 1986) cited in Hoover (1995, p. 6).
The counterfactual analysis discussed above will be employed in Chapter 7
in order to assess the macroeconomic management ability of the Lao
Government. To avoid the “Lucas critique” the alternative assumptions for
policy experiments will be made by allowing the policy variables to vary
along the path of their historical trends. By doing this, it is arguable that the
policy experiments will not be subject to the “Lucas critique”, as changes in
policy variables will not change the economic environment faced by
economic agents.
Chapter 4__________________________________________________________________ Page 103
4.5 Data sources and variable constructions
Apart from the Asian Development Bank (ADB), other international
organisations like the World Bank, IMF, and United Nations have published
macroeconomic data needed for the analysis in this study. However, data on
national accounts reported by these organisations often differ and some
observations are also not available (see Table A.4.1.1). To ensure that data
sets are consistent, it is desirable to avoid using data from many sources.
Thus, in this study the data for most of the macroeconomic variables are
from the Key Indicators of Developing Asian and Pacific Countries, which is
published annually by the ADB. Regarding data on aid inflows to the Lao
PDR, the World Bank and United Nations have published this information.
However, some observations are not available from the World Bank source.
To have consistent data, this study only used aid flow data published
annually by the United Nations. To convert all variables to 1990 constant
price, a GDP deflator and the average exchange rate index are used as the
deflator variables. This is due to the lack of time-series data for import prices
and export prices as well as the consumer price index (CPI). Data for the
period 1978 to 1997 is used for the estimation of the macroeconometric
models, while that for 1998 to 2001 is used for the counterfactual analysis.
It should be note that data on income (Y) for the years 1978 to 1979 are not
available from the ADB Source. Therefore, data have been obtained from
the Bank of Lao PDR. Also, there is a lack of consistent data for the private
sector, mainly from 1988 backwards. Variables are therefore constructed for
private saving, private investment, private consumption, and disposable
income.
To construct the private saving variable, the saving equation used by the
World Bank (1994) is employed as follows.
SPY = −0.2337 + 0.0508 ln(YPC ) + 0.1485(QMY )
(4.49)
Chapter 4__________________________________________________________________ Page 104
where SPY is the ratio of private saving to GDP, ln YPC is the natural log of
per capita income, and QMY is the ratio of quasi-money to GDP. The ratio of
private saving to GDP can be constructed using the available data on YPC
and QMY and substituting into equation (4.49). In the next step, the national
accounting identity is utilised to derive private investment variable as
follows:
(S − I ) = ( X − M )
(4.50)
(SP + SG) − (IP + IG ) = ( X − M )
(4.51)
IP = (SP + SG) − IG − ( X − M )
(4.52)
where data on SG, IG, X and M are available from both the fiscal and balance
of payments accounts. Once both private savings and private investment
variable are constructed, private consumption (CP) and disposable income
(YD) can be constructed using the national accounting identity as follows:
YD = Y − T + TR
(4.53)
CP = YD − SP
(4.54)
where Y is income, T is government revenue, and TR is transfer payments.
The construction for the growth rate of effective labour force is based on the
study by Gounder and Xayavong (2001). The effective labour force is
defined as follows:
LE t = SE t
GLE t =
LFt
POPt
LE t − LE t −1
LE t −1
(4.55)
(4.56)
where LEt is the effective labour force, SEt is the total number of student
enrolments in secondary and tertiary levels, LFt is labour force, POPt is total
population, and GLEt is the growth rate of the effective labour force. Note that
the LFt variable is employed because of the lack of employment data in the
Chapter 4__________________________________________________________________ Page 105
Lao PDR. There will be no distinction between the use of labour force and
employment if unemployment is constant or varies within a small range.
The real effective exchange rate index (REER) is constructed using the
trade-weighted real exchange rate index between the US dollar and the Thai
baht. The geometric averaging technique is used. The formula of REER
expressed in term of foreign currency is given by:
E
REER =  US
 PUS
WUS




 ETH

P
 TH
WTH




PLAO
(4.57)
where the subscript letter US, TH and LAO respectively represents the
United States of America, Thailand and the Lao PDR, E is the index of the
nominal exchange rate defined as units of foreign exchange per unit of
domestic currency, P is aggregate price index, W is the trade-weighted index
for multiple trading partners. The sum of trade-weighted index must be equal
to one (i.e., WUS + WTH = 1 ). An upward movement in the real effective
exchange rate index indicates an appreciation of the REER.
4.6 Summary
This chapter has presented the macroeconometric methods utilised for
analysing the aid-growth nexus. It also has explained the methods of data
construction employed in this study. The Keynesian approach to modelling
macroeconometric system is selected to solve the existing methodological
weakness of aid fungibility and aid-growth models. Unlike the conventional
aid fungibility model, which focuses only to the government sector, the
macroeconometric model of aid fungibility developed in this chapter allows
aid to have direct, indirect and feedback effects on public, private and trade
sectors.
Whether or not a stable aid inflow contributes to economic growth depends
on the potential effect of aid on investment through the savings-investment
channel. In other words, aid fungibility may or may not displace private
Chapter 4__________________________________________________________________ Page 106
investment which depends on whether aid inflow results in an increase in
domestic borrowing. Unlike the typical approach of using a single regression
equation, the macroeconometric models of the aid-growth nexus allows the
tracking of the dynamic impact of capital inflows on economic growth through
the potential effects of aid on investments along the external balance
channel. Multiplier analysis is introduced to capture the interactive effect of
aid and policy conditionality on economic growth. Counterfactual analysis is
also introduced to disentangle the effects of the weakness of the aid
recipient’s state and institutional capability on economic growth from the
effect of aid. The issues and methodology discussed here are employed in
the next chapters.
Chapter 4__________________________________________________________________ Page 107
Appendix 4.1: Sources and the estimation results of macroeconomic
data
This Appendix consists of three tables. Table A4.1.1 presents data
comparison between the author’s estimation, and the World Bank, the ADB,
and the IMF. Table A4.1.2 presents data used for the estimation of the aid
fungibility model in Chapter 5, and Table A4.1.3 presents data used for the
estimation of the aid-growth nexus model.
Table A4.1.1 Comparing Estimated Data with Other Sources
Year
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
IMF(2002)
Y
n.a.
n.a.
355868
410429
429780
442674
471175
514154
538983
533810
522600
574221
612681
637100
681700
712800
780600
835600
893200
955000
Author's estimation
IY
SY
8.81
-10.18
6.84
-0.92
10.20
-1.53
9.37
1.79
17.44
4.84
12.67
4.06
11.90
4.40
13.89
3.83
12.17
6.39
12.29
7.56
24.77
2.76
19.23
2.59
15.20
2.87
14.85
3.66
17.37
5.73
22.12
7.69
25.64
8.56
25.20
9.55
30.59
10.97
28.70
12.16
World bank (1999)
Y
IY
SY
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
6.18
2.60
n.a.
7.04
1.31
n.a.
6.93
2.84
n.a.
10.19
2.26
624456
13.50
-4.77
713682
n.a.
n.a.
761277
n.a.
n.a.
792718
n.a.
n.a.
847024
n.a.
n.a.
897104
n.a.
n.a.
970299
n.a.
n.a.
1038275
26.00
11.57
1109730
30.50
11.97
1182304
28.67
11.35
ADB (various issues)
Y
IY
SY
425600*
n.a.
n.a.
429800*
n.a.
n.a.
435080
n.a.
n.a.
449850
n.a.
n.a.
459850
n.a.
n.a.
469500
n.a.
n.a.
480500
n.a.
n.a.
497234
n.a.
n.a.
521342
n.a.
n.a.
515750
n.a.
n.a.
506278
14.9
2.2
574172
15.1
1.1
612681
14.8
0.8
637160
15.9
4.6
681798
14.5
5.6
721819
15.2
6.5
780657
n.a.
n.a.
835519
24.5
11.5
893256
29.0
12.4
955009
26.2
9.4
Legend: IY is gross domestic investment as percentage of GDP, SY is gross domestic savings as percentage of GDP,
Y is income at constant market price 1990, n.a. indicates data are not available, * data obtained from the Bank of Lao
PDR.
Notes: The differences of data on income (Y) reported by the above-mentioned sources may be caused by differences
in data compilation, and also the use of different deflation to convert of income to constant prices. With respect to data
on gross domestic investment (IY) and gross domestic savings (SY), the author’s estimation is not much different from
other sources’ during the year 1990 to 1997. However, from 1989 backward the author’s estimation is higher than other
sources’.
Chapter 4__________________________________________________________________ Page 108
Table A4.1.2: Data used for the estimation of the aid fungibility
macroeconometric model
Year
Y
CP
CG
IG
IP
X
M
MK
1978
4256
4088
693
340
35
128
949
231
1979
4298
4174
368
226
67
228
828
219
1980
4351
4217
546
397
46
304
996
345
1981
4499
4246
406
366
56
233
1104
385
1982
4599
4311
524
746
56
383
1267
569
1983
4695
4432
419
534
61
380
1394
505
1984
4805
4499
462
477
95
397
1468
408
1985
4972
4749
557
600
91
444
1601
505
1986
5213
4962
621
492
143
419
1413
437
1987
5158
4835
514
432
201
387
1302
285
1988
5063
4459
621
1041
213
525
1353
389
1989
5742
5170
532
885
219
508
1555
646
1990
6127
5485
699
736
195
562
1324
622
1991
6372
5594
723
610
336
654
1151
842
1992
6818
5899
746
664
520
763
1553
929
1993
7218
6035
818
724
872
1311
2353
1162
1994
7807
6404
914
909
1093
1520
2854
1318
1995
8355
6827
902
959
1147
1473
2791
1256
1996
1997
8933
9550
7251
7716
893
878
1030
1080
1702
1661
1554
1513
3323
3085
1562
1776
Average
Std. Deviation
5941
1648
5268
1099
642
176
662
257
441
546
684
491
1683
756
719
463
Year
YD
T
DT
NT
A
FDI
PDF
REER
1978
4156
191
100
91
920
0
0.6
168.4
1979
4253
251
45
205
637
0
1.1
294.6
1980
4299
397
52
345
440
0
1.9
207.7
1981
4342
390
156
234
353
0
2.5
386.8
1982
4419
639
180
459
367
0
4.3
255.6
1983
4545
497
150
347
278
0
7.0
503.3
1984
4618
554
187
367
310
0
8.9
632.2
1985
4890
607
82
524
307
0
17.0
455.3
1986
5140
776
74
702
367
0
23.8
589.3
1987
5094
645
63
581
351
0
31.2
658.9
1988
4587
632
476
156
667
18
45.2
72.2
1989
5377
473
365
108
1163
32
75.1
100.5
1990
5750
610
376
233
1080
43
100.0
100.0
1991
5892
659
480
179
970
54
113.3
108.0
1992
6305
730
513
218
949
52
123.8
112.9
1993
6526
881
692
189
1124
360
131.7
116.0
1994
7005
981
801
179
1102
305
141.9
118.7
1995
7508
1019
847
171
1446
451
171.2
120.8
1996
1997
8004
8621
1120
1135
929
929
191
206
1604
1671
769
422
193.2
230.4
115.3
131.0
Average
Std. Deviation
5567
1351
659
266
375
316
284
166
805
459
125
216
71
74
262
200
Legend: Y is income, the letter P stands for the private sector and letter G stands for the government sector, C is
consumption, I is investment, X is total exports, M is total imports, MK is imports of capital, YD is disposable income, T
is total government revenue, DT is tax revenue, NT is non-tax revenue, A is total aid inflow, FDI is total foreign direct
investment, PDF is GDP deflator, REER is real effective exchange rate.
Note: all variables unit are100 Million Kips at 1990 prices, except the PDF and REER units which are in percentages.
Source: ADB (various), UN (Various) and author’s estimation.
Chapter 4__________________________________________________________________ Page 109
Table A4.1.3: Data used for the estimation of the aid-growth nexus
macroeconometric model
Year
GY
IY
IGY
IPY
XY
MY
MKY
MCY
1978
0.90
8.81
7.99
0.82
3.01
22.29
5.42
16.87
1979
1980
1981
0.99
1.23
3.39
6.84
10.20
9.37
5.27
9.13
8.14
1.57
1.06
1.24
5.32
7.00
5.18
19.26
22.90
24.55
5.10
7.94
8.57
14.16
14.96
15.99
1982
1983
2.22
2.10
17.44
12.67
16.22
11.36
1.22
1.31
8.33
8.10
27.54
29.68
12.38
10.75
15.17
18.93
1984
1985
2.34
3.48
11.90
13.89
9.92
12.06
1.98
1.83
8.26
8.93
30.55
32.20
8.49
10.15
22.06
22.05
1986
1987
1988
4.85
-1.07
-1.84
12.17
12.29
24.77
9.44
8.38
20.56
2.74
3.91
4.21
8.03
7.51
10.36
27.11
25.25
26.73
8.38
5.52
7.68
18.73
19.72
19.06
1989
1990
13.41
6.71
19.23
15.20
15.41
12.01
3.82
3.19
8.85
9.17
27.09
21.61
11.25
10.15
15.84
11.46
1991
1992
4.00
7.01
14.85
17.37
9.57
9.74
5.28
7.63
10.27
11.19
18.07
22.78
13.21
13.62
4.86
9.16
1993
1994
1995
5.87
8.15
7.03
22.12
25.64
25.20
10.03
11.64
11.47
12.09
14.00
13.73
18.16
19.48
17.63
32.60
36.57
33.41
16.10
16.88
15.03
16.51
19.68
18.37
1996
1997
6.91
6.91
30.59
28.70
11.53
11.31
19.06
17.39
17.39
15.84
37.20
32.30
17.48
18.60
19.72
13.70
Average
Std. Deviation
4.23
3.61
16.96
7.03
11.06
3.32
5.90
5.92
10.40
4.74
27.48
5.49
11.14
4.15
16.35
4.27
Year
SY
TY
AY
FDIY
REER
PDF
GLE
YF
1978
1979
-10.18
-0.92
4.48
5.83
21.63
14.82
0.00
0.00
168.4
294.6
0.6
1.1
3.80
5.68
1.90
2.06
1980
1981
-1.53
1.79
9.13
8.68
10.12
7.85
0.00
0.00
207.7
386.8
1.9
2.5
9.87
14.50
2.31
3.63
1982
1983
4.84
4.06
13.89
10.59
7.98
5.91
0.00
0.00
255.6
503.3
4.3
7.0
4.74
5.27
3.24
3.52
1984
1985
1986
4.40
3.83
6.39
11.52
12.20
14.89
6.45
6.17
7.04
0.00
0.00
0.00
632.2
455.3
589.3
8.9
17.0
23.8
-2.63
9.65
2.35
5.23
3.04
3.79
1987
1988
7.56
2.76
12.50
12.48
6.80
13.17
0.00
0.36
658.9
72.2
31.2
45.2
-0.31
-0.12
6.18
8.55
1989
2.59
8.24
20.25
0.56
100.5
75.1
-1.93
6.07
1990
2.87
9.95
17.62
0.70
100.0
100.0
-1.72
6.25
1991
3.66
10.34
15.23
0.85
108.0
113.3
-3.75
5.26
1992
1993
5.73
7.69
10.71
12.20
13.92
15.58
0.76
4.99
112.9
116.0
123.8
131.7
12.33
2.23
5.33
5.25
1994
8.56
12.56
14.12
3.90
118.7
141.9
12.32
6.51
1995
9.55
12.19
17.31
5.39
120.8
171.2
6.87
6.46
1996
1997
10.97
12.16
12.54
11.88
17.95
17.50
8.61
4.42
115.3
131.0
193.2
230.4
5.26
-32.54
5.03
-0.08
Average
Std. Deviation
4.34
4.95
10.84
2.57
12.87
5.14
1.53
2.50
262.4
200.1
71.2
73.9
2.59
9.84
4.48
2.05
Legend: GY is output growth, the letter Y is attached to the following variable to indicate the measurement of that
variable as a percent of GDP: IY is total investment, IGY is public investment, IPY is private investment, XY is total
exports, MY is total imports, MKY is imports of capital, MCY is imports of consumption goods, SY is gross domestic
savings, TY is total government revenue, AY is total aid inflow, FDIY is total foreign direct investment, REER is real
effective exchange rate, PDF is GDP deflator, GLE is the growth rate of the effective labour force, YF is the weighted
average by trade share of growth in per capita income of the Lao PDR’s trading partners.
Note: all variables are in percentage units.
Source: ADB (various), UN (Various) and author’s estimation.
Chapter 5__________________________________________________________________ Page 110
Chapter 5
AID FUNGIBILITY AND ITS EFFECTS ON INVESTMENT
…Foreign aid in different times and different places has thus
been highly effective, totally ineffective, and everything in
between… The evidence, however, is that aid is often fungible,
so that what you see is not what you get…
World Bank (1998, p. 2 & p. 5)
There is nothing inherently wrong or inappropriate about
fungibility; all it indicates is that donors and recipients have
differing views about how expenditures should be allocated.
The presumption that donors are right need not always hold.
McGillivray and Morrissey (2000, p. 419)
5.1 Introduction
It has been pointed out in Chapters 2 and 3 that a stable aid inflow
contributes to economic growth even when aid is fungible, which may imply
that aid fungibility posed no concerns for aid allocation. However, the
implications of aid fungibility remain a topic of vigorous debate. According to
the World Bank policy research report, Assessing Aid (World Bank, 1998),
foreign aid is largely fungible, indicating that recipient governments are
effectively able to avoid donor attempts to target aid flows to specific sectors.
Assessing Aid reinforces the point that donors should not consider giving aid
to governments that are not committed to sectors for which this aid is
targeted, as “where donors and governments do not agree on the allocation
of expenditure, …spending is not likely to be effective” (World Bank, 1998,
p. 61). This assertion, however, has been challenged on many aspects.
Lensink and White (2000, p. 11) criticised such suggestive strategy for aid
allocations because while in the donors’ view this may improve the quality of
aid and limit sectoral aid fungibility, however it can over-state the effect of
aid fungibility. McGillivray and Morrissey (2000) argue that aid fungibility
distracts attention from the true issues of fiscal management. Instead of
targeting the allocation of foreign aid to a specific sector, they suggest that
Chapter 5__________________________________________________________________ Page 111
more attention should be paid to “how aid impacts on public sector
behaviour overall and how fiscal management can be improved” (ibid., p.
413).
There is no doubt that both sides have plausible arguments for the
improvement of aid effectiveness. As discussed in the displacement theories
in Chapter 2, aid fungibility may lower the effectiveness of aid if aid is
allocated to low productive sectors and/or an increase in aid inflow leads to
lowering tax revenue. The latter effect may crowd out private investment
through the need to raise Public Sector Borrowing Requirement (PSBR). It is
made clear in Assessing Aid that effective aid does not replace private
initiative but acts as a magnet and crowds in private investment (World
Bank, 1998, p. 3).
To address the issue of aid fungibility mentioned above, this chapter
employs the macroeconometric model of aid fungibility to analyse the impact
of aid fungibility on investments in the case of the Lao PDR. Before
presenting the macroeconometric analysis, Section 5.2 considers the
potential impacts of aid and policy conditionality on investment in the Lao
PDR. The descriptive analysis is used to assess the potential impact of aid
on investment within the saving-investment channel. Section 5.3 reports the
estimation results of the behavioural equations and the validity test for the
model and also presents the empirical evidence of the aid fungibilityinvestment nexus. Finally, Section 5.4 provides the conclusion and policy
implications.
5.2 Foreign aid, policies and the internal balance
According to the displacement theory discussed in Chapter 2, a high level of
aid inflow has a potential side effect on investment via the internal balance
(i.e., saving-investment channel). In other words, a high level of aid inflow
tends to reduce tax efforts, lower domestic savings, increase government
consumption, and crowd out private investments. If this is the case, a surge
Chapter 5__________________________________________________________________ Page 112
in foreign aid will not raise investment by as much as the value of the foreign
aid inflows. This argument, however, seems to have over-stated the true
problems facing the Lao PDR’s recent development efforts. An extremely low
tax base and weak government institutional capacity limited the ability of the
Lao Government to mobilise domestic resources to supplement aid inflow.
The discussions in the “fiscal response” literature also indicate that how the
potential side effect of an increase in aid inflow affects tax and saving efforts
depends on the way in which the aid-receiving government conducts fiscal
policy.
Following the implementation of economic reforms in 1989, the Lao
Government has been progressive in policy and institutional reforms. These
improvements have created favourable environments for supporting
domestic resource mobilisation.
Figure 5.1: Aid and government’s revenues in the Lao PDR, 1985-2001
25.0%
20.0%
15.0%
10.0%
5.0%
Government revenue as % of GDP
2001
2000
1999
1998
1997
1996
1995
1994
1993
1992
1991
1990
1989
1988
1987
1986
1985
0.0%
Aid as % of GDP
Source: Data extracted from Appendix 5.1, Table A5.1.1.
As illustrated in Figure 5.1, aid increased in the 1988-89 year, two years
after the launching of the “adjustment programmes” in 1986. The level of aid
inflow as a percentage of GDP was almost tripled from the annual average of
about 6 percent in the 1985-88 period to about 16 percent of GDP after
1989. This upsurge in aid inflow has a negative relationship with the
government’s revenues. However, this does not mean that the increased
inflow of foreign aid lowered tax revenue. In fact, tax collections in the Lao
Chapter 5__________________________________________________________________ Page 113
PDR are affected by the nature of the tax base and the lack of institutions.
The latter has resulted in the J-curve effect (Bourdet, 2000, pp. 73-75). For
example, the introduction of a new tax system in 1988 caused a fall in
revenue in the following year. After 1990, revenue, however, rebounded and
increased steadily to an average of just less than 12 percent of GDP for the
1990-96 period. In addition, although the Lao Government has taken
measures to improve the domestic tax system, “the tax base remains narrow
with taxes on foreign trade being the largest revenue contributor” (Lao
Government, 2001, p. 10). The heavy dependence on trade taxes was also
the main reason used to explain the revenue fluctuation after the Asian
financial crisis in 1997, as the decline in external trade and the use of an
outdated and overvalued exchange rate to calculate import tariffs resulted in
lowered trade tax collections (ADB, 2000, p. 96). The fluctuation in tax
collections were attributed to the chronic lack of transparency in the
transactions of state and business sectors (Bourdet, 2000, p. 103).
2001
2000
1999
1998
1997
1996
1995
1994
1993
1992
1991
1990
1989
1988
1987
1986
1985
Figure 5.2: Aid and savings in the Lao PDR, 1985-2001
25.0%
20.0%
15.0%
10.0%
5.0%
0.0%
-5.0%
Aid as % of GDP
Gross domestic saving as % of GDP
Government saving as % of GDP
Source: Data extracted from Appendix 5.1, Table A5.1.1.
In Figure 5.2, an increase in aid inflow has been initially associated with the
decline in gross domestic saving for the 1988-1992 period. However, this
does not necessarily imply that a high level of aid inflow always reduces
savings. Under the “adjustment programmes”, the Lao Government was
required to stimulate savings in the public sector. As public saving is itself a
Chapter 5__________________________________________________________________ Page 114
component of gross domestic saving (GDS), this policy directly raised GDS.
For example, the level of public saving swung from negative in the 1989-92
period to positive in the 1993-97 period, which led to an escalation in GDS
after 1992. The level of GDS as a percentage of GDP doubled from the
annual average of about 5.4 percent for the 1985-92 period to about 9.7
percent for the 1993-1997 period. With the surplus of public saving, the Lao
Government was able to increase wages in order to replace the food coupon
payments to public officials. The Lao Government’s adoption of a large wage
increase led to an increased level of expenditure on government
consumption. Note that the surge of GDS was associated with the increase
in private saving that was attributable to income growth during the 1993-97
period. However, the fall in savings in 1999 was attributable to the volatile
macroeconomic environment that took place after the Asian financial crisis.
To encourage domestic savings, the government undertook financial reforms
to improve competitiveness in the banking sector. The government also
permitted foreign currency deposits that allowed depositors, who are
uncertain about the value of local currency, to avoid foreign exchange risk
while holding deposits in domestic banks. This development helped to bring
back the capital that was previously held abroad and thus contributed to
increased financial deepening, in which most of the total deposits come from
increased foreign currency deposits.1 As illustrated in Figure 5.3 below, the
degree of financial deepening (i.e., the ratio of deposits including foreign
currency to GDP) rapidly increased from less than 5 percent in 1986 to
almost 20 percent in 1999. The fall in the level of financial deepening during
the 1995-96 period was attributed to the enforcing of an administrative
restriction on foreign exchange and the crackdown in the parallel exchange
market in 1995. This move shook the public’s confidence in the local
currency and caused capital flight.
1
The Bank of Thailand reported that there were individual Lao deposit accounts of about $US 5.45 million in 1991 and
$US 6.54 million in 1992 deposited in Thai border banks. These amounts equal approximately two thirds of total private
sector current deposits, savings deposits and foreign currency deposits in the Lao commercial banks at that time (Bank
of Thailand, 1992).
Chapter 5__________________________________________________________________ Page 115
To bring back public’s confidence, the government adopted a freely floating
exchange rate system and lifted all outstanding exchange rate restrictions.
This move contributed to increased financial deepening over the 1997-99
period. The decline in domestic currency deposits was attributable to
depositors switching their accounts denominated in the domestic currency to
the US dollar (IMF, 2000). From this observation, it can be put forward that
high aid inflows would not displace domestic savings as long as the Lao
Government creates an environment conducive to savings.
Figure 5.3: Financial deepening in the Lao PDR, 1985-2001
2001
2000
1999
1998
1997
1996
1995
1994
1993
1992
1991
1990
1989
1988
1987
1986
1985
20.0%
18.0%
16.0%
14.0%
12.0%
10.0%
8.0%
6.0%
4.0%
2.0%
0.0%
M2 as % of GDP (inclusive of foreign currency deposits)
M2 as % of GDP (exclusive of foreign currency deposits)
Source: Data extracted from Appendix 5.1, Table A5.1.1.
It should be noted that although domestic savings have increased, the level
of saving is still far behind the financing needed to satisfy a rapid growth in
domestic investment demand. This is due to the fact that the country is
extremely poor. “A survey conduced in 1995 by the World Bank in selected
provinces indicated that the Lao PDR had approximately 46 percent and 53
percent of the total and rural population, respectively, living in poverty with
per capita income of less than US$100 per annum” (Lao Government, 2001,
p. 4).
With respect to the potential impact of foreign aid on private investment, the
possibility that increased public spending may crowd out private investment
probably depends on how much the public sector borrowing requirement
Chapter 5__________________________________________________________________ Page 116
(PSBR) changes to supplement aid inflows. Under donors’ conditionality,
measures to reduce budget deficits and domestic borrowing are often the
norm; and in the case of the Lao PDR, most of the public investment
programmes are financed by aid inflows. Therefore, this policy tends to
reduce the PSBR and leaves the available domestic resources for expanding
investment in the private sector. In other words, aid flows are not likely to
crowd out private investment.
Figure 5.4: Foreign aid, private investment and public investment in the
Lao PDR, 1985-2001
25.0%
20.0%
15.0%
10.0%
5.0%
Private investment as % of GDP
Aid as % of GDP
2001
2000
1999
1998
1997
1996
1995
1994
1993
1992
1991
1990
1989
1988
1987
1986
1985
0.0%
Public investment as % of GDP
Source: Data extracted from Appendix 5.1, Table A5.1.1.
As illustrated in Figure 5.4, the correlation between the flow of foreign aid,
public and private investment appears to be positive for most periods. The
level of public investment as a percentage of GDP reached its peak at 20.6
percent in 1988, before declining an average annual rate of about 10
percent of GDP over the 1989-97 period. Approximately 80 percent of the
Public Investment Programme (PIP) was financed through foreign aid of
which 70 percent of the PIP was allocated to the construction of road,
electricity distribution and irrigation networks (Lao Government, 1997). The
level of private investment also increased from 1.8 percent of GDP in 1985
to about 22 percent of GDP in 1997. This rapid growth of private investment,
especially after 1993, was due to the high level of FDI inflow.
Chapter 5__________________________________________________________________ Page 117
However, in the wake of the Asian financial crisis, the sharp drop in private
investment from its peak of 17.5 percent of GDP in 1996 to about 7.3 percent
of GDP in 1998 and 8.3 percent of GDP in 1999 was related to the decline in
FDI both in value and in number of projects (see for example IMF, 2000,
Table 26 and 27). The decline in aid flows since 1997 was related to the IMF
and other aid donors withholding financial aid. The ratio of foreign aid to
GDP for the 1997-99 period declined by almost 50 percent from its peak
level of 18 percent in 1996. The indigenous private sector was also hit hard
by this crisis, as many large firms dependent on FDI and government
contracts were financed by foreign aid (MPDF, 1997, p. 37). On the other
hand, public investment rose more than 30 percent above its peak level in
1997. The increase in public investment was related to the Lao
Government’s increased spending on ambitious investment in irrigation
projects to promote self-reliance in rice production.
5.3 Estimating, validating the model and multiplier analysis of
aid fungibility and its impact on investment
The discussions in the preceding section seems to suggest that the potential
side effect of an increase in aid inflow on investment through the savinginvestment channel can be mitigated by carrying out appropriate
macroeconomic policies to stimulate tax reform and increase the mobilisation
of domestic savings. In other words, an increased aid inflow is unlikely to
lower saving and displaces private investment. To measure the magnitude of
aid fungibility and its impact on private investment, the macroeconometric
model of aid fungibility developed in Chapter 4 is employed here. It also will
be convenient to duplicate this model as follows:
Chapter 5__________________________________________________________________ Page 118
Macroeconometric model of aid fungibility
Behavioural Equations
Public Sector
CGt = f 1 (Tt , At , INFt )
IGt = f 2 (Tt , At , INFt )
(5.1)
(5.2)
Tt = f 3 (Yt , At , FDI t , INFt )
(5.3)
Private Sector
IPt = f 4 ( IG, YDt , FDI t , INFt )
CPt = f 5 ( YDt ; INFt )
(5.4)
(5.5)
Identity Equations
DI
t
=
f 1 ( DI
*
t
,Tt , At )
YDt = Yt + NTRt
ADt = ASt = Yt
(5.6)
(5.7)
(5.8)
where CG: government consumption spending,
IG: government capital expenditure,
T : government revenue,
IP: private investment,
CP: private consumption,
YD: disposable income,
Y: income,
NTR: net transfer payments,
A: total aid inflows,
FDI: foreign direct investment,
INF: inflation,
NX: trade balance,
AD: aggregate demand, and
AS: aggregate supply,
The behavioural equations are estimated using annual data covering the
1978-97 period. The estimation results and model validations are presented
below in Section 5.4.1. Finally, empirical evidence of the impact of aid
fungibility on investment is presented in Section 5.4.2.
Chapter 5__________________________________________________________________ Page 119
5.3.1 Estimation results and model validation
Table 5.1 below reports the results of testing for the long-run relationship
and the critical value bounds for each behavioural equation employed in the
macroeconometric model of aid fungibility. As the computed F-statistic for
each behavioural equation exceeds the upper bound of the critical values of
F-statistic, a conclusive decision can be made that variables in each
behavioural equation are cointergrated. Therefore, the ARDL methodology
can be applied to estimate the behavioural equations without requiring the
knowledge of whether the underlying variables employed in the behavioural
equations are I(0) or I(1).
Table 5.1: Results for Testing the Long-run Relationship of behavioural
equations in the macroeconometric model of aid fungibility: F-statistic
Models
EQ (5.1): CGt=f1(Tt; At, INFt)
EQ (5.2): IGt=f2 (Tt; At, INFt)
EQ (5.3): Tt=f3(Yt; At,FDIt, INFt)
EQ (5.4): IPt=f4(IGt, , YDt; FDIt, INFt)
EQ (5.5): CPt=f5(YDt; INFt)
Computed
F-statistic
4.537
Testing the existence of a long-run relationship:
critical values of the F-static
Significance level
F-Statistic case 2
3.182-4.126
90%
4.937
3.182-4.126
90%
5.299
2.711-3.800
90%
3.902
2.711-3.800
90%
26.674
4.042-4.788
90%
The estimation results of the behavioural equations are presented in Table
5.2 below. Overall, all behavioural equations have a relatively high
explanatory power in terms of R2. Some explanatory variables are retained in
each equation, despite the coefficients being statistically insignificant
because they improve the tracking ability of the simulation model. The
estimation results support the potential effects that aid may have on various
endogenous variables, as discussed in the preceding sections. Especially,
an increased aid inflow has not only a positive effect on both government
consumption spending (CG) and public investment (IG) but it also has a
negative effect on government revenue (T). Public investment, as expected,
has a positive effect on private investment (IP), while an increased inflow of
foreign direct investment (FDI) has a positive impact on government revenue
(T) and private investment (IP).
Chapter 5__________________________________________________________________ Page 120
Table 5.2: Estimation results of behavioural equations
Equation (5.1): Government consumption spending (CG)
CGt = 315.6 - 0.078CGt-1 + 0.619Tt - 0.229Tt-1+ 0.148At + 0.015At-1 - 0.920INFt
(104.0)* (0.200)
(0.150)* (0.171)
(0.094)
(0.101)
(0.733)
R2 =0.92, DW= 1.83, Durbin-h statistic =-0.72
Equation (5.2): Government Capital Expenditure (IG)
IGt = 264.5 + 0.038IGt-1 + 0.166Tt + 0.061Tt-1
(206.9 ) (0.223)
(0.403)
(0.487)
+ 0.828At - 0.616At-1 + 1.086INFt + 112.9D97
(0.237)* (0.261)* (1.814)
(198.6)
R2 =0.83, DW= 2.27
Equation (5.3): Government Revenue (T)
Tt = -669.2 + 0.290Tt-1 + 0.091Yt + 0.141Yt-1 - 0.307At + 0.003At-1
(559.9) (0.331)
(0.137) (0.210)
(0.135)*** (0.045)
+ 0.094FDIt - 0.257FDIt-1 +1.720INF
(0.600)
(0.569)
(0.147)
R2 =0.95, DW= 2.03
Equation (5.4): Private Investment (IP)
IPt = - 374.1 + 0.072IPt-1 + 0.133∆IGt + 0.110YDt + 0.003YDt-1
(296.3) (0.229)
(0.091)
(0.075)
(0.095)
+ 1.042FDIt + 0.650FDIt-1 - 1.217INFt
(0.139)*
(0.325)***
(0.928)
R2 =0.99, DW= 1.75
Equation (5.5): Private Consumption (CP)
CPt = 334.1 - 0.591CPt-1 + 0.859YDt – 0.536YDt-1 + 0.204INFt
(212.7) (0.234)**
(0.299)*
(0.181)*
(0.447)
R2 =0.99, DW= 1.97
Note: The number in parentheses under the coefficient is the corresponding
standard error. Statistical significance of a coefficient at the 1 %, 5%, and 10
% levels, respectively, are represented with *, **, and ***.
Chapter 5__________________________________________________________________ Page 121
Variable List:
Endogenous Variables consist of:
Exogenous and Dummy Variables:
CG: government consumption spending, AID: total aid inflows,
IG: government capital expenditure,
FDI: foreign direct investment,
T: government revenue,
INF: inflation,
IP: private investment,
D97 dummy variable for the
CP: private consumption,
Asian financial crisis takes the
YD: disposable income, and
value of 1 for years 1997
Y: income.
and 0 otherwise.
The model’s validation is carried out by performing an ex-post simulation, in
which a fully dynamic simulation process has been performed for the 198097 period in order to evaluate the ability of the model to track the historical
path of endogenous variables. As the initially proposed model structure
yields unstable dynamic behaviour, the behavioural equation of private
consumption is then treated as an exogenous variable.2 Exogenising the
private consumption variable does not affect the analysis of aid fungibility
because private consumption is little directly affected by the aid variable.
Figure 5.5 to Figure 5.10 plot the simulated values and the actual values of
seven endogenous variables, which show the model’s capability in tracking
turning points and trends in the endogenous variables.
2
For the method of tuning and adjusting simulation models, see for example, Pindyck and Rubinfeld (1991, pp. 390-93).
It should also be noted that the structure of a model might be such that its dynamic behaviour verges on being unstable.
As such, minor changes in the structure might be sufficient to stabilise the behaviour. This method is used throughout
this study to improve the simulation of the model.
Chapter 5__________________________________________________________________ Page 122
Figure 5.5: Simulated values of government consumption spending
(CGS) and actual value of government consumption spending (CG) at
constant 1990 prices (100 million Kips)
100Million Kips
1200
1000
800
600
400
200
CG
1997
1996
1995
1994
1993
1992
1991
1990
1989
1988
1987
1986
1985
1984
1983
1982
1981
1980
0
CGS
Source: Data extracted from Appendix 5.1, Table A5.1.2
1200
1000
800
600
400
200
IG
1997
1996
1995
1994
1993
1992
1991
1990
1989
1988
1987
1986
1985
1984
1983
1982
1981
0
1980
100Million Kips
Figure 5.6: Simulated values of government capital expenditure (IGS)
and actual values of government capital expenditure (IG) at constant
1990 prices (100 million Kips)
IGS
Source: Data extracted from Appendix 5.1, Table A5.1.2
100Million Kips
Figure 5.7: Simulated values of government revenue (TS) and actual
values of government revenue (T) at constant 1990 prices (100 million
Kips)
1500
1000
500
T
Source: Data extracted from Appendix 5.1, Table A5.1.2
TS
1997
1996
1995
1994
1993
1992
1991
1990
1989
1988
1987
1986
1985
1984
1983
1982
1981
1980
0
Chapter 5__________________________________________________________________ Page 123
Figure 5.8: Simulated values of private investment (IPS) and actual
values of private investment (IP) at constant 1990 prices (100 million
Kips)
1500
1000
500
IP
1997
1996
1995
1994
1993
1992
1991
1990
1989
1988
1987
1986
1985
1984
1983
1982
1981
0
1980
100Million Kips
2000
IPS
Source: Data extracted from Appendix 5.1, Table A5.1.2
Figure 5.9: Simulated values of disposable income (YDS) and actual
values of disposable income (YD) at constant 1990 prices (100 million
Kips)
100Million Kips
10000
8000
6000
4000
2000
YD
1997
1996
1995
1994
1993
1992
1991
1990
1989
1988
1987
1986
1985
1984
1983
1982
1981
1980
0
YDS
Source: Data extracted from Appendix 5.1, Table A5.1.2
Figure 5.10: Simulated values of income (YS) and actual values of
income (Y) at constant 1990 prices (100 million Kips)
100Million Kips
12000
10000
8000
6000
4000
2000
Y
Source: Data extracted from Appendix 5.1, Table A5.1.2
YS
1997
1996
1995
1994
1993
1992
1991
1990
1989
1988
1987
1986
1985
1984
1983
1982
1981
1980
0
Chapter 5__________________________________________________________________ Page 124
The summary of the various statistical tests, computed for the 1980-1997
period, is shown for each endogenous variable in Table 5.3. Some brief
comments about the model validation are given below.
Table 5.3: Statistical tests from the model validation
Endogenous
variable
CG
IG
T
IP
YD
Y
Actual
mean
661
723
726
510
5801
6231
Simulated
mean
674
725
754
527
5909
6338
SME
SMPE
RMSE
RMSPE
U
13.63
2.53
27.66
17.14
107.50
107.30
2.52
2.03
3.41
2.84
1.29
1.12
64.10
110.70
92.57
61.00
282.20
282.20
10.01
15.00
12.90
24.80
4.71
4.40
0.05
0.07
0.06
0.04
0.02
0.02
Legend: C is consumption, I is investment, the letter P stands for the private sector and letter G stands for the
government sector, T is government revenue, Y is income, YD is disposable income, SME is the simulation mean error,
SMPE is the simulation mean percent error, RMSE is the root mean square error, RMSPE is the root mean square
percent error and U is the Theil inequality coefficient.
Over all, the ex-post simulation indicates that the model can reproduce the
general movements of most actual data quite well. All the endogenous
variables have the Theil inequality coefficient (U) value almost equal to zero,
suggesting that the model for each endogenous variable has a relatively
high tracking ability. However, the model does not capture some of the
fluctuations in the public sector that occurred mostly in the pre-economic
reform period. The systematic bias statistic for each endogenous variable
measured by the simulation mean error (SME) and simulation mean percent
error (SMPE), indicates that the model is overestimated by 1.12 to 3.41
percent.3 It should be noted that the SME and SMPE might be close to zero if
large positive errors cancel large negative errors. In this regard, the root
mean simulation error (RMSE) and the root mean simulation percent error
(RMSPE) are used to measure the deviation of simulated variable from its
actual time path. The deviation of all variables is less than 15 percent,
except for the private investment variable (IP).
3
For a discussion of systematic bias see Pindyck and Rubinfeld (1991, p. 338). The positive (negative) values of SME
and SMPE suggest that the model is overestimated (underestimated).
Chapter 5__________________________________________________________________ Page 125
5.3.2 Empirical evidence of aid fungibility and its impact on investments
The impact of aid fungibility on investment is measured by aid fungibility
coefficients at the aggregate level or the multiplier effect of aid on the fiscal
variable (CG, IG, T). These coefficients are summarised in Table 5.4. The
figure in each column represents the multiplier effect on each endogenous
variable for every 1 unit increase in the exogenous variables. Some
comments about the impacts of aid on investment and income growth are
made below.
Table 5.4: Short run and long run multiplier effects of foreign aid
Endogenous
variable
CG
IG
T
IP
YD
Y
Short run multiplier
for aid variable
0.016
0.793
-0.214
0.220
1.030
1.030
Endogenous
variable
CG
IG
T
IP
YD
Y
Long rung multiplier
for aid variable
0.110
0.844
-0.077
0.134
1.088
1.088
Legend: C is consumption, I is investment, the letter P stands for the private sector and letter G stands for the
government sector, T is government revenue, Y is income, YD is disposable income.
As pointed out in Section 5.2, an increased aid inflow tends to raise both
government consumption spending (CG) and public investment (IG) and also
lower government revenue (T). This seems to suggest that aid is fungible. To
certify this result, the conditions of aid fungibility discussed in Chapter 2 are
applied as follows:
Short-run multiplier
Long-run multiplier
∆IG
= 0.793 < 1
∆A
∆IG
= 0.844 < 1
∆A
∆CG
= 0.016 > 0
∆A
∆CG
= 0.110 > 0
∆A
∆T
= −0.214 < 0
∆A
∆T
= −0.077 < 0
∆A
Chapter 5__________________________________________________________________ Page 126
The estimated multiplier values indicate that aid is fungible on three counts.
First, an increase in aid by 1 unit led to a short-run and long-run increase in
public investment by 0.793 and 0.844, respectively. This implies that aid
intended for public investment did not increase the value of public
investment by as much as the value of aid inflow. According to displacement
theory, this circumstance would imply that the Lao Government was able to
avoid donor attempts to target the allocation of aid for public investment
projects. Second, an increase in aid by 1 unit led to a short-run and long-run
increase in government consumption spending by 0.016 and 0.110 units,
respectively. Third, an increase in aid by 1 unit led to a short-run and longrun decrease in government revenues by 0.214 and 0.077 unit, respectively.
The last two counts would indicate that the Lao Government was able to use
resources released due to an increase in aid inflow, to increase consumption
or reduce tax revenue.
The multiplier analysis supports the preliminary examination carried out in
Section 5.2, which suggests that aid is fungible. Although aid is found to be
fungible, this does not imply that aid fungibility lowers the effectiveness of
aid. As explained in Section 5.2, an increase in spending on government
consumption improves the wellbeing of public officials. In addition, the
weakness in fiscal institutions and the extreme low income per capita of the
Lao PDR have resulted in low tax collections. Thus, it is not surprising to
observe an increase in aid associated with a decline in government revenue.
From this aspect, it can be argued that there was nothing inappropriate
about aid fungibility in the case of the Lao PDR. In fact, the commitments on
tax and institutional reforms that the Lao Government has made throughout
the period of the “adjustment programmes” have resulted in improved level
of tax collection. This improvement can be seen from the long-run increase
in the magnitude of government revenue with respect to an increase in aid
inflow (i.e., the multipliers of government revenues in response to aid were
increased from -0.214 in the short run to -0.077 in the long run).
Chapter 5__________________________________________________________________ Page 127
To clarify the puzzle as to whether aid fungibility issues lower the
effectiveness of aid, the analysis is extended to focus on the impact of aid
fungibility on private investment. This can be examined using the condition
of fiscal response to aid inflow proposed by White and McGillivray (1992), as
discussed in Chapter 2. By substituting the long-run multiplier of the fiscal
response to aid inflow as suggested in the White and McGillivray conditions,
the long-run impact of aid fungibility on private investment is expressed
mathematically as follows:
∆G ∆CG ∆IG
∆G
=
+
= 0.110 + 0.844 = 0.954 ⇒ 0 <
<1
∆A
∆A
∆A
∆A
(a)
∆T
∆T
= −0.077 ⇒
<0
∆A
∆A
(b)
∆G ∆T
−
= 0.954 − (− 0.077 ) = 1.031
∆A ∆A
(c)
⇒
∆G ∆T
∆PSBR
−
>1 ∴
>0
∆A ∆A
∆A
(d)
The conditions indicated in (a), (b) and (d) capture the Lao Government’s
fiscal behaviour in response to aid inflow. These three conditions are
consistent with case 8 expressed in Table 2.2, which implies that the Lao
Government’s fiscal policy may crowd out private investment. This is
because aid inflows, partly used to finance an increase in government
consumption expenditure and partly to fund tax reductions, led to an
increase in the public sector borrowing requirements (PSBR). However, as
the magnitude of the difference between the increase in government
spending and the decline in government revenue is almost equal to one (i.e.,
c ≈ 1), this implies that an increase in PSBR is almost equal to zero (i.e., d ≈
0). Thus, it can be said that conditions indicated in (a), (b) and (d) are
consistent with case 9 expressed in Table 2.2, which indicates no crowding
out effect on private investment. As explained in Section 5.2, an increase in
Chapter 5__________________________________________________________________ Page 128
aid inflow is unlikely to crowd out private investment. This indication is
supported by the aid multipliers, by which an increase in aid inflow indirectly
raises private investment (IP) through the direct effect of an increase in
public expenditure (i.e., the variable capturing the crowding effect of fiscal
policy). As illustrated in Table 5.4 above, an increase of 1 unit in the aid
inflow raises private investment to 0.220 units in the short run and 0.134
units in the long run.
Based on the above discussion, it can be concluded that an increase in aid
did raise both public and private investment, then it should be expected that
aid might contribute to economic growth although aid is fungible in the case
of the Lao PDR. Indeed, as illustrated in Table 5.4, aid has a positive effect
on income, i.e., every 1 unit increase in aid inflow leads to a short-run and
long-run increase in income by 1.030 and 1.088 units, respectively.
However, the aid fungibility model developed in this chapter is a demanddriven model, in which the trade sector and the aggregate supply side are
treated as exogenous variables. Thus, this interpretation of the impact of aid
on growth should be taken with caution because the foreign exchange
constraints on either investment or capacity utilisation have not been
incorporated into the supply side of the economy.
5.4 Conclusion
This chapter has examined the issue of aid fungibility and its impact on
investment in the case of the Lao PDR, using descriptive analysis and a
demand-driven macroeconometric model. The findings indicate that though
aid was fungible, it did not crowd out private investment. The explanation for
this outcome is that an increased aid inflow was associated with the nonincrease in the public sector borrowing requirement. This in turn reflects the
effectiveness of donors’ conditionality on fiscal policy. In this regard, an
increase in aid inflow leaves available resources for expanding investment in
the private sector. However, the appearance of aggregate aid fungibility is
attributed to the lack of domestic resources to supplement aid inflows. The
Chapter 5__________________________________________________________________ Page 129
extreme low income and the weakness of fiscal institutions have made it
difficult to implement policies to mobilise domestic resources in the Lao
PDR.
From the above point, it would be wrong to use aid fungibility as a way to
exclude poor countries from being aid recipients. Instead of targeting the
allocation of aid to a specific sector, donors should support poor countries to
improve their policies and institutional capacity. Indeed, the analysis in this
chapter implies that policy conditionality has mitigated the side effects of aid
on investments in the saving-investment channel. The analysis also reveals
that a stable aid inflow has enabled the Lao Government to raise investment.
This finding is likely to indicate that an increase in aid inflow will boost
economic growth through the increase in demand for consumption and
investment in both public and private sectors. However, drawing a very
confident conclusion about the potential impact of stable aid inflow on
economic growth requires the further examination of the potential impact of
aid on investment in the supply side model. This subject will be undertaken
in Chapter 6.
Chapter 5__________________________________________________________________ Page 130
Appendix 5.1: Selected macroeconomic indicators, and simulation
results of aid fungibility model
This appendix consists of two tables. Table A5.1.1 presents some
macroeconomic indicators. Table A5.1.2 presents the simulation results of
aid fungibility model.
Table A5.1.1: Selected macroeconomic indicators: the Lao PDR, 19852001
Year
1985
1986
1987
1988
1989
1990
1991
1992
1993
Gross domestic saving as % of GDP
3.8%
6.4%
7.6%
2.8%
2.6%
2.9%
3.7%
5.7%
7.7%
Government saving as % of GDP
Government revenue as % of GDP
Public investment as % of GDP
1.0% 3.0% 2.5% 0.2% -1.0% -1.5% -1.0% -0.2% 0.9%
12.2% 14.9% 12.5% 12.5% 8.2% 9.9% 10.3% 10.7% 12.2%
12.1% 9.4% 8.4% 20.6% 15.4% 12.0% 9.6% 9.7% 10.0%
Private investment as % of GDP
Aid as % of GDP
1.8%
6.2%
2.7%
7.0%
3.9%
6.8%
4.2% 3.8% 3.2% 5.3% 7.6% 12.1%
13.2% 20.3% 17.6% 15.2% 13.9% 15.6%
M2 as % of GDP (inclusive of foreign currency deposits)
M2 as % of GDP (exclusive of foreign currency deposits)
2.7%
1.5%
3.1%
1.6%
5.3%
5.0%
5.8%
5.2%
4.9%
4.4%
7.2%
4.2%
7.1%
4.5%
9.1%
5.3%
13.2%
8.7%
1997
1998
1999
2000
2001
Year
1994
1995
1996
Gross domestic saving as % of GDP
Government saving as % of GDP
8.6%
0.9%
9.6%
1.4%
11.0% 11.5% 15.5% 13.4% 12.7% 14.6%
2.5% 2.7% 0.3% 2.3% 4.4% 5.0%
Government revenue as % of GDP
Public investment as % of GDP
Private investment as % of GDP
12.6% 12.2% 12.5% 10.9% 9.8% 11.3% 13.2% 13.5%
11.6% 11.5% 11.5% 11.3% 15.5% 14.4% 12.8% 13.3%
14.0% 13.7% 19.1% 17.4% 7.3% 8.3% 6.9% 7.5%
Aid as % of GDP
M2 as % of GDP (inclusive of foreign currency deposits)
M2 as % of GDP (exclusive of foreign currency deposits)
14.1% 17.3% 18.0% 17.5% 14.2% 15.2% 13.2% 13.9%
15.0% 13.5% 14.2% 18.3% 18.4% 18.7% 17.5% 18.2%
9.7% 8.0% 8.5% 7.9% 7.3% 6.9% 4.2% 4.6%
Source: ADB (various issues), UN (various issues), IMF (various issues) and Author’s estimation.
Chapter 5__________________________________________________________________ Page 131
Table A5.1.2: Simulation results of aid fungibility model: 1990-1997
Year
1980
1981
1982
1983
1984
1985
1986
1987
1988
CG
IG
TS
IP
YD
Y
CGS
IGS
TS
IPS
YDS
YS
Year
546
397
397
46
4299
4351
495
414
420
55
4436
4488
1989
406
366
390
56
4342
4499
464
428
406
72
4182
4338
1990
524
746
639
56
4419
4599
482
552
488
54
4335
4515
1991
419
534
497
61
4545
4695
489
478
539
30
4266
4416
1992
462
477
554
95
4618
4805
475
511
477
91
4318
4505
1993
557
600
607
91
4890
4972
510
577
604
60
4657
4739
1994
621
492
776
143
5140
5213
541
585
616
173
5193
5267
1995
514
432
645
201
5094
5158
582
537
680
182
5159
5222
1996
621
1041
632
213
4587
5063
559
815
616
179
4708
5184
1997
CG
IG
T
IP
YD
Y
532
885
473
219
5377
5742
699
736
610
195
5750
6127
723
610
659
336
5892
6372
746
664
730
520
6305
6818
818
724
881
872
6526
7218
914
909
981
1093
7005
7807
902
959
1019
1147
7508
8355
893
1030
1120
1702
8004
8933
878
1080
1135
1661
8621
9550
CGS
IGS
TS
IPS
YDS
YS
596
1051
553
280
5685
6050
696
663
656
304
6010
6386
741
613
777
439
6410
6890
782
684
896
492
6555
7068
826
862
971
867
6856
7548
818
752
994
996
6835
7636
893
1073
1060
1283
7911
8758
1004
1026
1214
1739
8322
9251
1002
1121
1272
1711
9049
9978
Legend: C is consumption, I is investment, T is total revenue government revenues, YD is disposable income, Y is
income. The letter P stands for the private sector, letter G stands for the government sector, and letter S stand for
simulation.
Note: all variables unit are100 Million Kips of 1990 prices.
Chapter 6 __________________________________________________________________Page 132
Chapter 6
THE EFFECTS OF A STABLE AID FLOW
ON ECONOMIC GROWTH
…We have at best an imperfect understanding of the
determinants of growth and perhaps less still of how aid affects
those determinants... If it could be demonstrated that aid
increases investment or imports... then most economists would
accept that these are channels through which aid contributes to
higher growth...
White (1992b, p. 133)
…Rather than embarking on a rather fruitless debate on
whether or not economic reforms work by comparing
differential outcomes between ESAF and non-ESAF countries,
it is now more important to understand the mechanisms
through which programmes do, or do not, work, and to assess
why programmes have had more positive outcomes in some
countries than in others, and at certain times rather than others,
and whether positive outcomes are sustainable.
UNCTAD (2000a, p. xii)
6.1 Introduction
It has been emphasised in Chapter 3 that it is necessary for aid-receiving
countries to mobilise both domestic and external resources, as this effort will
enable them to overcome resource and foreign-exchange constraints on
capital accumulation and economic growth. Nevertheless, the analysis in
Chapter 5 has indicated that the Lao PDR could not raise enough domestic
resources and thus foreign capital inflows remained the main resource
dominating capital accumulation and economic growth in the Lao PDR. By
the same token, it was also expected that a stable aid inflow coupled with
the Lao Government’s commitment to liberalise the economic system to help
boost the demand side of the economy should have created favourable
conditions for economic growth.
Chapter 6 __________________________________________________________________Page 133
To examine the effects of stable aid inflows on economic growth in the Lao
PDR, the analysis carried out in Chapter 5 is further extended in this chapter
in order to take into account any external resource constraints on the supply
side of the Lao PDR’s economy. The analysis focuses on the interactive
effects between a stable aid flow and policy prescriptions on economic
growth through the investment channel. The macroeconometric model of the
aid-growth nexus developed in Chapter 4 is employed here. Before
presenting the macroeconometric analysis, Section 6.2 highlights the
potential impacts of aid and policy conditionality on the Lao PDR’s ability to
raise investment. Whether a high level of aid inflow displaces or
supplements investment in the export-import channel is closely examined.
Section 6.3 presents the results of the estimation and validation of the
macroeconometric model. The empirical results of the interactive effects
between aid and donors’ conditionality on economic growth are also
discussed. Section 6.4 presents the conclusion and policy implication.
6.2 Foreign aid, policies and external balances
The discussions in the “Dutch disease” literature indicate that aid may lower
export earnings (ability to import and ability to raise investment) via the side
effect of a high level of aid inflow on the recipient country’s external
competitiveness. In other words, if an increased aid inflow leads to an
appreciation in the real effective exchange rate (REER), aid will not raise
investment as much as the value of aid inflow. This argument, however,
seems to be inadequate for explaining the export performance of the Lao
PDR. One of the crucial factors influencing the export performance of the
Lao PDR was the improvement in the economic environment bought about
by the Lao Government’s commitment to undertake the “adjustment
programmes”. A rapid economic liberalisation and the policies to promote
foreign direct investment (FDI) led to greater efficiency of resource uses and
a substantial capital inflow. Otani and Pham also add that:
Chapter 6 __________________________________________________________________Page 134
...price and trade liberalisation reduced the costs of imported
inputs and eliminated monopoly rents, and widespread
privatisation enhanced productivity in the industrial sector. In
addition, technological change may have favoured the
production of tradables rather than nontradables, as the former
benefited from foreign direct investment, competition, and
economies of scale.
Otani and Pham (1996, p. 55)
Certainly, if the benefits derived from the “adjustment programmes” are
substantial, this tends to outweigh the negative effect of the real effective
exchange rate appreciation on export performance. Indeed, as Otani and
Pham point out, during the 1988-94 period of a high-level aid and FDI inflow,
although the REER appreciated by either about 60 percent (measured by
purchasing power parity) or by about 20 percent (measured by terms of
trade), export performances remained strong (Otani and Pham, 1996, pp.
52-53). Figure 6.1 below also indicate that the share of exports to GDP grew
from about 10 percent during the 1988-92 period to about 20 percent during
the 1993-94 period.
Figure 6.1: Aid flows and exports in the Lao PDR, 1985-2001
25.0%
20.0%
15.0%
10.0%
5.0%
Aid as % of GDP
Total exports as % of GDP
2001
2000
1999
1998
1997
1996
1995
1994
1993
1992
1991
1990
1989
1988
1987
1986
1985
0.0%
Foreign Direct investment as % of GDP
Source: Data extracted from Appendix 6.1 Table A6.1.1.
The lack of diversity in its export commodities and its reliance on a few
select markets also influenced the Lao PDR’s export performance. As
illustrated in Table 6.1 below, three groups dominate the composition of
Chapter 6 __________________________________________________________________Page 135
export commodities: electricity exports, agricultural and manufacturing
products. Although in the early 1990s FDI inflow was followed by an
increase in the share of manufactured products in total exports, the export
performance of the Lao PDR remained influenced by the changes in market
conditions of the Lao PDR’s main trading partners.
Table 6.1: Composition of the Lao PDR’s exports, 1988-1999
Year
1, Agricultural Products
2, Manufacturing Products
3, Electricity
4, Other
1988
73%
0%
19%
8%
1989
56%
6%
24%
14%
1990
51%
9%
24%
15%
1991
49%
29%
22%
0%
1992
40%
43%
13%
5%
1993
33%
51%
8%
8%
Year
1, Agricultural Products
2, Manufacturing Products
3, Electricity
4, Other
1994
37%
47%
8%
8%
1995
39%
44%
8%
9%
1996
52%
33%
9%
6%
1997
40%
39%
7%
15%
1998
32%
31%
20%
17%
1999
33%
32%
26%
9%
Source: ADB (1991, 1994, 1997, 2001), and IMF (1996, 1997, 1998a, 2000).
For example, the decline in export performance in 1995 was attributed to the
decline in the production of wood products and the demand for motorcycles
that were exported to Thailand and Vietnam. The loss of the general special
preference (GSP) status of garment exports to the EU coupled with a low
level of FDI inflow to the tradable sector caused a decline in the performance
of the export sector after 1996. Although the Lao PDR regained GSP status
in 1997, the performance of the export sector remained broadly unchanged
from earlier years, as export demand had been affected by the financial
crisis in Thailand. Electricity exports were revitalised by a substantial inflow
of FDI into the hydropower industry, and the rise in the level of electricity
exports as a percentage of GDP in 1998-99 was due to a new
hydroelectricity plant starting transmission. A recovery in wood exports, and
Chapter 6 __________________________________________________________________Page 136
the increased penetration of garments into the European market also
contributed to the recovery of the export sector during the 1998-99 period.1
So far, the above discussions are likely to suggest that the export
performance in the Lao PDR was influenced by not only the appreciation of
REER but also by the changes in demands for the Lao PDR’s export
commodities. Therefore, caution is needed when evaluating the potential
effect of high level aid inflow on export performance in the Lao PDR. In fact,
the analysis in Chapter 5 indicates that aid inflow did not displace
investment in the saving-investment channel. The above analysis also
indicates that aid did not lower export earnings. In this regard, it is expected
that aid would also not displace investment in the export-import channel.
Figure 6.2: Aid flows and imports in the Lao PDR, 1985-2001
Capital imports as % of Total imports
Aid as % of GDP
2001
2000
1999
1998
1997
1996
1995
1994
1993
1992
1991
1990
1989
1988
1987
1986
1985
80.0%
70.0%
60.0%
50.0%
40.0%
30.0%
20.0%
10.0%
0.0%
Total imports as % of GDP
Source: Data extracted from Appendix 6.1, Table A6.1.1
As illustrated in Figure 6.2, although the share of total import to GDP was
declining during the 1989-91 period, an increase in aid inflow after 1989
enabled the Lao PDR to raise the level of capital goods. The explanation for
this circumstance is that a large component of aid inflow has always been
allocated to the public investment programmes (PIP) that are often linked to
imports of capital goods. Thus, the share of imported capital goods in total
1
For more details of discussion about export performances in the Lao PDR see IMF (1996, 1997, 1998a, 2000).
Chapter 6 __________________________________________________________________Page 137
imports rapidly increased and reached its peak (73 percent) in 1991. Overall,
as aid inflow remained high and stable during the 1991-97 period, the
shares of capital goods in total imports rose from an annual average of
about 30.9 percent during the 1985-89 period to about 53.1 percent during
the 1990-97 period.
It should be noted that although an increase in aid inflow enabled the Lao
PDR to raise the level of imported capital goods, the structure of imports had
a strong bias towards consumption goods (i.e., the share of capital imports is
less than 50 percent for most of the time). Thus, apart from the influence of
aid, the trends in total imports were dominated by the changes in the imports
of consumption goods. For example, the introduction of stabilisation
programmes during the 1989-91 period required the Lao Government to
reduce the level of budget deficits and control money growth. These policy
measures led to a decline in imports of consumption goods and hence the
share of total imports. Again, the decline in the share of total import since
1996 was related to the Lao Government imposing import restrictions on
consumption goods, and also the weak demand for imported intermediate
and capital goods linked to the financial crisis in Thailand and the fall in the
investment demands of the hydropower industry. Except for 1997, the rise in
capital imports was related to the increased public investment in irrigation
projects.
6.3 Estimating, validating the model and multiplier analysis of
aid, economic growth and policies
The analysis in the preceding section seems to suggest that the “adjustment
programmes” attached to aid has mitigated the side effect of a high level of
aid inflow on investment in the export-import channel. In fact, an increased
aid inflow has enabled the Lao PDR to raise capital imports and investment.
Thus, it is expected that a high level aid inflow should make a contribution to
economic growth via investment channel. To see this, the macroeconometric
Chapter 6 __________________________________________________________________Page 138
model of the aid-growth nexus developed in Chapter 4 is employed to
measure the interactive effect of aid and policy conditionality on investment
and economic growth. Before presenting the analyses, it will be convenient
to duplicate this model in this section as follows:
A specification of the aid-growth nexus model
Behavioural Equations
Output Growth
GYt = f1 ( IYt ; GLEt )
(6.1)
Investment
IGYt = f 2 (GYt ; AYt , INFt )
(6.2)
IPYt = f 3 ( GYt , IGYt ; FDIYt , INFt )
(6.3)
International Trade
XYt = f 4 ( MKYt ; YFt REERt )
MKYt = f 5 ( IYt ; REERt )
(6.4)
(6.5)
Identity Equations
IYt = IGYt + IPYt
MYt = MKYt + MCYt
(6.6)
(6.7)
CAYt = XYt − MYt + U t
(6.8)
where GY:
IY:
IGY:
IPY:
XY:
MKY:
MCY:
MY:
CAY:
AY:
FDIY:
INF:
GLE:
YF:
growth rate of income,
the share of investment in GDP,
the share of public investment in GDP,
the share of private investment in GDP,
the share of exports to GDP,
the share of imports of capital goods in GDP,
the share of imports of consumption goods in GDP,
the share of imports to GDP,
the ratio of current account deficit to GDP,
the ratio of total aid to GDP,
the ratio of total foreign direct investment to GDP
the inflation rate,
the growth rate of effective labour force,
the weighted average by trade share of growth in per
Chapter 6 __________________________________________________________________Page 139
capita income of the trading partners,
REER: real exchange rate,
U:
the error term including net service transfers.
The behavioural equations are estimated using annual data covering the
1978-97 period. The same estimation method employed in Chapter 5 has
been applied here. The estimation results and model validations are
presented below in Section 6.4.1 and the empirical evidence of the aidgrowth nexus is presented in Section 6.4.2.
6.3.1 Estimation results and model validation
Table 6.3 below reports the results from testing the long-run relationship,
and the critical value bounds for each behavioural equation employed in the
macroeconometric model of aid fungibility. As the computed F-statistic for
each model exceeds the upper bound of the critical values of F-statistic, a
conclusive decision can be made that variables in each behavioural
equation are cointergrated. Therefore, the ARDL methodology can be
applied to estimate the behavioural equations without requiring the
knowledge of whether the underlying variables employed in the behavioural
equations are I(0) or I(1).
Table 6.2: Results for testing the long-run relationship of behavioural
equations in a macroeconometric model of the aid-growth nexus: Fstatistic
Models
EQ (6.1): GYt=f1(IYt; GLEt)
EQ (6.2): IGYt=f2(GYt; AYt, INFt)
EQ (6.3): IPYt=f3(GYt; IGYt,FDIYt, INFt)
EQ (6.4): XYt=f4(MKYt,; YFt, REER t)
EQ (6.5): MKYt=f5(IYt; REERt)
Computed
F-statistic
Testing the existence of a long-run relationship:
critical values of the F-statistic
Significance level
F-Statistic case1
3.4091
2.458-3.342
90%
5.4283
2.180-3.211
90%
3.7856
2.022-3.112
90%
3.3341
2.180-3.211
90%
6.9148
2.458-3.342
90%
Table 6.3 below reports the results of the estimation of behavioural
equations. Overall, the behavioural equations have a relatively high
explanatory power in terms of R2, except the public investment equation that
has R2 less than 0.8. Some explanatory variables are retained in each
Chapter 6 __________________________________________________________________Page 140
equation, despite statistically insignificant coefficients because they improve
the tracking ability of the simulation model. All coefficients of behavioural
equation have the expected signs.
Table 6.3: Estimation results of behavioural equations
Equation (6.1): Output Growth (GY)
GYt = -0.719 - 0.021GYt-1 - 0.048IYt + 0.441IYt-1
(2.021) (0.102)** (0.089)
(0.100)*
+ 0.071GLEt + 0.0008GLEt-1 - 6.370DWEA
(0.039)***
(0.068)
(1.419)*
R2 =0.81, DW= 1.97, Durbin-h statistic =0.00
Equation (6.2): Public Investment (IGY)
IGYt = 8.892 - 0.021IGYt-1 - 0.157∆GYt + 0.942AYt - 0.839AYt-1 + 0.039INFt
(2.027)* (0.167)
(0.131)
(0.201)* (0.161)*
(0.015)*
R2 =0.74, DW= 1.57
Equation (6.3): Private Investment (IPY)
IPYt = 3.151 + 0.427IPYt + 0.036∆IGYt + 0.050∆GYt
(0.831)* (0.206)*** (0.055)
(0.054)
+ 1.044FDIYt + 0.238FDIYt-1 - 0.031INFt
(0.187)*
(0.355)
(0.010)**
R2 =0.98, DW= 1.6, Durbin-h statistic =1.79
Equation (6.4): Total export (XY)
XYt = -11.81 + 0.164XYt-1 + 0.991MKYt + 0.315MKYt-1 + 0.006REERt-1+ 0.829YFt
(3.11)* (0.194)
(0.182)*
(0.235)
(0.002)**
(0.208)*
R2 =0.95, DW= 2.09
Equation (6.5): Imported capital goods (MKY)
MKYt = 4.8327 + 0.7019MKYt + 0.1768IYt - 0.1352IYt-1 - 0.0065REERt-1
(0.888)* (0.087)*
(0.082)** (0.085)
(0.001)*
R2 =0.92, DW= 2.39
_____________________________________________________________
Note: The number in parentheses under the coefficient is the corresponding
standard error. Statistical significance of a coefficient at the 1 %, 5%, and 10
% levels, respectively, are represented with *, **, and ***.
Chapter 6 __________________________________________________________________Page 141
Variable list:
Endogenous variables:
Exogenous variables:
GY: growth rate of real GDP,
IGY: share of public investment in GDP,
IPY: share of private investment in GDP,
XY: share of export in GDP,
MKY: share of imported capital goods
in GDP,
IY: share of total investment in GDP,
MY: share of total import in GDP,
CAY: ratio of current account deficits
to GDP
AY: ration of aid inflows to GDP,
FDIY: ratio of foreign direct
investment to GDP
INF: inflation,
D97: dummy variable for the
Asian financial crisis takes
the value of 1 for 1997
and 0 otherwise,
DWEA: dummy variable for
weather conditions takes
the value of 1 for
years 1987 and 1988
and 0 otherwise.
The dynamic simulation process has been conducted for the 1980-97 period
for evaluating the tracking ability of the model. Figure 6.3 through to Figure
6.10 plot the simulated values against the actual values of all endogenous
variables employed to evaluate the tracking ability of the model. The statistic
test of the model validation is reported in Table 6.4. Some comments on the
tracking ability of the model are made below.
15.00
10.00
(%) 5.00
0.00
-5.00
GY
Source: Data extracted from Appendix 6.1, Table A6.1.2.
GYS
1997
1996
1995
1994
1993
1992
1991
1990
1989
1988
1987
1986
1985
1984
1983
1982
1981
1980
Figure 6.3: Simulated values of real output growth (GYS) and actual
values of real output growth (GY)
Chapter 6 __________________________________________________________________Page 142
Figure 6.4: Simulated values of public investment (IGYS) and actual
values of public investment (IGY)
25.00
20.00
15.00
(%)
10.00
5.00
IGY
1997
1996
1995
1994
1993
1992
1991
1990
1989
1988
1987
1986
1985
1984
1983
1982
1981
1980
0.00
IGYS
Source: Data extracted from Appendix 6.1, Table A6.1.2.
Figure 6.5: Simulated values of private investment (IPYS) and actual
values of private investment (IPY)
25.00
20.00
(%)
15.00
10.00
5.00
IPY
1997
1996
1995
1994
1993
1992
1991
1990
1989
1988
1987
1986
1985
1984
1983
1982
1981
1980
0.00
IPYS
Source: Data extracted from Appendix 6.1, Table A6.1.2.
Figure 6.6: Simulated values of total Investment (IYS) and actual values
of total investment (IY)
IY
Source: Data extracted from Appendix 6.1, Table 6.1.2.
IYS
1997
1996
1995
1994
1993
1992
1991
1990
1989
1988
1987
1986
1985
1984
1983
1982
1981
1980
35.00
30.00
25.00
20.00
(%)
15.00
10.00
5.00
0.00
Chapter 6 __________________________________________________________________Page 143
Figure 6.7: Simulated values of capital imports (MKYS) and actual
values of capital imports (MKY)
20.00
15.00
(%) 10.00
5.00
MKY
1997
1996
1995
1994
1993
1992
1991
1990
1989
1988
1987
1986
1985
1984
1983
1982
1981
1980
0.00
MKYS
Source: Data extracted from Appendix 6.1, Table 6.1.2.
Figure 6.8: Simulated values of total imports (MYS) and actual values of
total imports (MY)
MY
1997
1996
1995
1994
1993
1992
1991
1990
1989
1988
1987
1986
1985
1984
1983
1982
1981
1980
40.00
35.00
30.00
25.00
(%) 20.00
15.00
10.00
5.00
0.00
MYS
Source: Data extracted from Appendix 6.1, Table 6.1.2.
Figure 6.9: Simulated values of total exports (XYS) and actual values of
total exports (XY)
25.00
20.00
15.00
(%)
10.00
5.00
XY
Source: Data extracted from Appendix 6.1, Table 6.1.2.
XYS
1997
1996
1995
1994
1993
1992
1991
1990
1989
1988
1987
1986
1985
1984
1983
1982
1981
1980
0.00
Chapter 6 __________________________________________________________________Page 144
Figure 6.10: Simulated values of current account deficits (CAYS) and
actual values of current account deficits (CAY)
1997
1996
1995
1994
1993
1992
1991
1990
1989
1988
1987
1986
1985
1984
1983
1982
1981
-5.00
1980
0.00
-10.00
(%)
-15.00
-20.00
-25.00
CAY
CAYS
Source: Data extracted from Appendix 6.1, Table 6.1.2.
Table 6.4: Statistical test of the model validation
Endogenous
variable
GY
IGY
IPY
XY
MKY
IY
MY
CAY
Actual
mean
4.40
6.17
11.22
17.39
11.44
10.79
27.76
-16.97
Simulated
mean
4.37
6.21
11.26
17.47
11.47
10.25
27.80
-17.54
SME
SMPE
RMSE
RMSPE
U
-0.04
0.04
0.03
0.08
0.04
-0.54
0.04
-0.58
26.60
12.81
3.64
2.73
3.02
-8.03
0.44
3.67
1.94
0.83
2.16
2.20
1.32
2.02
1.32
2.20
68.61
41.79
20.86
13.51
17.15
27.48
5.19
15.17
0.18
0.05
0.09
0.06
0.05
0.09
0.02
0.06
Legend: GY is a growth of output, IGY is the share of public investment to GDP, IPY is the share of private investment to
GDP, MKY is the share of imports of intermediate and capital good to GDP, XY is the share of total export to GDP, IY is
the share of total investment to GDP, MY is the share of total import to GDP, CAY is the ratio of current account deficits
to GDP, SME is the simulation mean error, SMPE is the simulation mean percent error, RMSE is the root mean square
error, RMPSE is the root mean square percent error and U is the Theil inequality coefficient.
Overall, the model has captured most of the turning points of all historical
endogenous variables. This implies that the model has good tracking ability,
which is important for the measure of dynamic multipliers. However, the
simulated growth of real output (GYS) and the simulated public investment
variables (IGYS) appear to deviate from its historical trend. This is borne out
by the values of two statistics, in which the simulated growth rate of real
output deviated from its actual value by 26.6 percent (measured by SMPE)
and 68.61 percent (measured by RMSPE). This result may be attributable to
poor estimation of the public investment function which is the main variable
entering the real output growth function. The simulated public investment
variable deviated from its actual value by 12.81 percent (measured by SMPE)
Chapter 6 __________________________________________________________________Page 145
and 41.79 percent (measured by RMSPE). The deviation of the public
investment variable might be due to inaccurate reporting of public spending.
However, other system bias indicators (i.e. SME and U) for the growth rate of
real output and the public investment variables are close to zero. Therefore,
the model is still considered to have a good tracking ability for these two
variables.
6.3.2 Empirical evidence of the interactive effect of stable aid flow and
policy conditionality on economic growth
As mentioned earlier, completing the hard-core conditionality on policy and
institutional reforms under the “adjustment programmes” is the prerequisite
for the disbursement of aid. In this context, the aid multipliers can be used as
a proxy to capture the interactive effect of aid and policy conditionality on
economic growth.2 To measure this interactive effect, the dynamic multiplier
values for all endogenous variables (i.e. GY, IGY, IPY, MKY, XY, IY, MY, CAY)
are computed with respect to the ratio of aid inflow to GDP (AY). The
dynamic multiplier values of other variables that are expected to dominate
the Lao Government’s ability to raise domestic investment with external
viability are also computed. Those variables are the ratio of foreign direct
investment to GDP (FDIY), the real effective exchange rate (REER) and the
growth rate income of the trading partners (YF). The estimation results of the
short-run and long-run multipliers are presented in Table 6.5 below. The
figures in each column represent the multiplier effect on the endogenous
variables for every 1 unit increase of the exogenous variables. Note that the
aid multiplier value reveals a positive impact on private investment, implying
that aid did not crowd out private investment. Therefore, the estimated result
of the supply model is consistent with the estimated result of the demanddriven model. Some comments about the interactive effects between stable
aid flow and policy conditionality on economic growth are made below.
2
This assumption is based on the view that aid is delivered upon the satisfactory performance of the “adjustment
programmes”. A stable aid inflow implies significant changes in policy and institutional reforms have been made in the
Lao PDR.
Chapter 6 __________________________________________________________________Page 146
Table 6.5: The Short Run and Long Run Multiplier Effects of Foreign
Aid for the 1978-97 period.
Endogenous
variables
GY
IGY
IPY
XY
MKY
IY
MY
CAY
Endogenous
variables
GY
IGY
IPY
XY
MKY
IY
MY
CAY
the short run multiplier of some exogenous variables
AY
FDIY
REERt-1
YF
-0.047
-0.051
0.000
0.000
0.950
0.008
0.000
0.000
0.032
1.042
0.000
0.000
0.172
0.184
0.000
0.829
0.174
0.186
-0.007
0.000
0.982
1.050
0.000
0.000
0.174
0.186
-0.007
0.000
-0.001
-0.002
0.007
0.829
the long run multiplier of some exogenous variables
AY
FDIY
REERt-1
YF
0.302
0.597
0.000
0.000
0.923
0.000
0.000
0.000
0.000
1.825
0.000
0.000
0.201
0.398
-0.026
0.993
0.129
0.255
-0.022
0.000
0.923
1.825
0.000
0.000
0.129
0.255
-0.022
0.000
0.073
0.144
-0.004
0.993
Legend: GY is the growth of output, IGY is the share of public investment to GDP, IPY is the share of private investment
to GDP, MKY is the share of imports of intermediate and capital good to GDP, XY is the share of total export to GDP, IY
is the share of total investment to GDP, MY is the share of total import to GDP, CAY is the ratio of current account
deficits to GDP, AY is the ratio of aid to GDP, FDIY is the ratio of foreign direct investment to GDP, YF is the weighted
average by trade share of growth in per capita income of the Lao PDR’s trading partners, and REERt-1 is the lag of the
real effective exchange rate.
Although the interactive effect of aid and policy conditionality on economic
growth has been approached from different theoretical and methodological
aspects, this issue remains somewhat controversial in the aid effectiveness
literature. For example, Connors (1979), Pastor (1987) and Gylfason (1987)
reported no effect of the “adjustment programmes” on economic growth;
Killick (1995) found ambiguous effects of the “adjustment programmes”, and
Conway (1994) reported a negative effect on growth in the first year of the
“adjustment programmes” but the negative effects diminish thereafter. The
finding in the case of the Lao PDR supports the latter case, in which the
impact of aid on economic growth reflects the tradeoffs between the shortrun impact of stabilisation policies and the long-run goal of structural
adjustment policies to promote economic growth. As illustrated in Table 6.5,
in the short run the aid multiplier value (AY) shows a negative impact on
Chapter 6 __________________________________________________________________Page 147
output growth (GY); however, in the long run the multiplier value obtained is
positive. This outcome can be explained below.
Under the stabilisation programmes the Lao Government was encouraged to
pursue demand-reducing policies, which included cutting government
expenditure, raising taxation to reduce budget deficits, raising interest rates
and restraining domestic credit to improve the current account balance.
Although the potential effects of these policy measures on economic growth
are not explicitly included in the model, such policy measures have been
found in other studies to have a negative impact on economic growth. For
example, Khan and Knight (1985), Heller et al. (1988) and IMF (1998b) find
that an austere fiscal policy often causes a reduction in private income and
social spending, and thus reduces economic growth. Blejer and Cheasty
(1989) also finds that tightening monetary policy by raising real interest rates
can induce good firms to shut down along with bad ones, and can lead to a
reduction in economic growth. This effect is embedded in the short-run
multiplier of aid with respect to output growth.
On the effect of structural adjustment programmes on economic growth,
Heller et al. (1988, p. 16) suggest that a negative impact of stabilisation
policies in the short run may be offset in the long run by positive
consequences from supply-side policies to promote economic growth.
Indeed, the positive value of the aid multiplier for the case of the Lao PDR
indicates that the “adjustment programmes” attached aid can create a
favourable economic environment (i.e. a low inflation, a small deviation of
the official exchange rate from the market rate) for economic growth. Indeed,
during 1989-91, the success of the Lao Government in implementing the
stabilisation
environment.
programme
This
brought
favourable
about
economic
a
stable
environment
macroeconomic
coupled
with
preliminary success in implementing the structural adjustment programmes
contributed to a stable aid inflow and a substantial increase in investment by
the private sector (including FDI). Therefore, implementing the structural
Chapter 6 __________________________________________________________________Page 148
adjustment policies created a strong supply-side economy and also
accelerated economic growth during the 1990-97 period.3
It should be noted that the estimated magnitude of the aid-growth coefficient
found in Assessing Aid indicates that increasing aid by 1 percent of GDP
(AY) in countries with sound management policies translates to a sustained
increase in growth of 0.5 percentage points of GDP (World Bank, 1998, p.
14). However, the magnitude of the aid-growth coefficient found in the case
of the Lao PDR is slightly less than the one found in Assessing Aid, where
an increase in stable aid inflow by 1 percent of GDP (AY) in the Lao PDR
during 1989-97 period led to the long-run increase in economic growth (GY)
of 0.302 percent. This result is consistent with the study by Gomanee et al
(2002, p. 20), by which each one percentage point increase in the aid/GNP
ratio adds one-third of one percentage point to the growth rate of economic
growth in the case of sub-Saharan African Countries (SSA).
Now turning to the Lao PDR’s ability to raise investment with external
viability, the discussion in Section 6.2 suggests that a high level aid inflow
attached to the “adjustment programmes” is unlikely to displace investment
in the export-import channel. This indication is consistent with the estimated
multiplier values of aid with respect to exports, capital imports and
investment. For the export channel, an increase in aid by 1 percent of GDP
(AY) led to a short-run and long-run increase in the share of export to GDP
by 0.172 and 0.201 percent, respectively. It is obvious that the effect of aid
on exports outweighs the negative effect of the REER on the share of exports
to GDP (i.e., the multiplier values for the share of exports to GDP with
respect to the REER are equal to 0.000 in the short run, and –0.026 in the
long run, which is smaller than the effect of aid on the share of export to
GDP). For the import channel, increasing aid by 1 percent of GDP (AY) to
3
For further discussions of the potential impact of stabilisation and structural-adjustment policies see for example
Haggard, et al., (1995), IMF (1997b). Otani and Pham (1996) also discuss outcomes of the “adjustment programmes” in
the case of The Lao PDR.
Chapter 6 __________________________________________________________________Page 149
finance capital imports led to a short-run and long-run increase in the share
of capital imports to GDP (MKY) by 0.174 and 0.129 percent, respectively.
Aid-financed capital imports in turn raised the share of total investment to
GDP to 0.982 percent in the short run and 0.923 percent in the long run.
Therefore, in the case of the Lao PDR it can be concluded that aid attached
to the “adjustment programmes” contributed to economic growth via the
investment channel.
The discussion in Section 6.2 also suggests that apart from foreign aid, FDI
and the changes in demand for and market access conditions of the Lao
PDR’s export commodities are a crucial factor that influences export
performance. This indication is seen by the estimated positive multiplier
values of FDIY and YF. For example, an increase in FDI inflow by 1 percent
of GDP (FDIY) led to a short-run and long-run increase in the share of export
to GDP (XY) by 0.184 and 0.398 percent, respectively. Also, a 1 percent
increase in the growth of per capita income of the Lao PDR’s trading
partners (YF) raised the share of exports to GDP by 0.829 percent in the
short run and by 0.993 percent in the long run, (note that the market access
constraints are embedded in this multiplier). It is obvious that the magnitude
of FDIY and YF multiplier values with respect to XY are larger than the AY
multiplier values with respect to XY. Besides, it should be noted that FDI
made a contribution to export growth though its influences on the ability to
raise investment and capital good imports (i.e., the channel through which
FDIY affects XY is via Equations 6.5 to 6.7. Equation 6.6 also indicates that
the per capita income growth of the Lao PDR’s trading partners has direct
influences on export growth. This implies that both FDI inflow and the growth
in the Lao PDR’s trading partners’ economies are crucial factors that
influence the Lao PDR’s ability to raise investment with external viability. For
example, although in the short run an increase in FDI inflow by 1 percent of
GDP (FDIY) worsened the current account deficit by 0.002 percent, in the
long run it improved the current account deficit by 0.144 percent. Also, a 1
percent increase in the growth of per capita income of the Lao PDR’s trading
Chapter 6 __________________________________________________________________Page 150
partners (YF) improved the current account deficit by 0.829 and by 0.993
percent of GDP in the short and long run, respectively.
6.4 Conclusion
This chapter has presented the analysis of the aid-growth nexus in the case
of the Lao PDR, using descriptive analysis and the supply-side drivenmacroeconometric model. The findings indicate that the interactive effect of
aid and policy conditionality has a short-run and long-run trade-off between
the effects of stabilisation and structural adjustment programmes on
economic growth. In other words, aid attached to stabilisation programmes is
found to have a negative effect on economic growth in the short run. This
short-run outcome reflects the negative effect of demand-reducing policy
measures on private income and social spending. Thus, aid attached to
stabilisation programmes has retarded economic growth.
With respect to the effects of the structural adjustment programmes on
economic growth, after the post-stabilisation period a stable aid flow has
been found to have a positive effect on growth, whereby increased aid has
contributed to economic growth via the investment channel. Other factors
dominating the Lao PDR’s ability to raise investment are FDI and export
growth. An increase in FDI inflows coupled with strong export growth have
released the foreign exchange constraints on capital imports, thus enabling
the Lao PDR to raise investment with external viability. However, this
favourable condition has been interrupted by the Asian financial crisis of
1997 and the problem of policy mismanagement of the Lao Government.
These adverse effects resulted in the Lao PDR’s economy leading into a
recession, and have put much pressure on the Lao Government to search
for corrective policies in order to steer the economy back to the sustainable
growth path. Chapter 7 will further explore this issue in detail.
Chapter 6 __________________________________________________________________Page 151
Appendix 6.1: Selected Macroeconomic indicators and simulation
results of the aid-growth nexus model.
This appendix consists of two tables. Table A6.1.1 presents some
macroeconomic indicators. Table A6.1.2 presents the simulation results of
the aid-growth nexus model.
Table A6.1.1: Selected Macroeconomic indicators: the Lao PDR, 19852001
Year
1985
1986
1987
1988
1989
1990
1991
1992
1993
Total exports as % of GDP
8.9%
8.0%
7.5%
10.4%
8.8%
9.2%
10.3%
11.2%
18.2%
Total imports as % of GDP
32.2%
27.1%
25.2%
26.7%
27.1%
21.6%
18.1%
22.8%
32.6%
Capital imports as % of Total imports
31.5%
30.9%
21.9%
28.7%
41.5%
47.0%
73.1%
59.8%
49.4%
Foreign Direct investment as % of GDP
0.0%
0.0%
0.0%
0.4%
0.6%
0.7%
0.9%
0.8%
5.0%
Aid as % of GDP
6.2%
7.0%
6.8%
13.2%
20.3%
17.6%
15.2%
13.9%
15.6%
1994
1995
1996
1997
1998
1999
2000
2001
Year
Total exports as % of GDP
19.5%
17.6%
17.4%
15.8%
17.1%
15.5%
18.0%
18.5%
Total imports as % of GDP
36.6%
33.4%
37.2%
32.3%
27.7%
26.2%
28.0%
30.0%
Capital imports as % of Total imports
46.2%
45.0%
47.0%
57.6%
40.9%
45.0%
60.7%
63.3%
Foreign Direct investment as % of GDP
3.9%
5.4%
8.6%
4.4%
2.3%
3.9%
5.9%
6.4%
Aid as % of GDP
14.1%
17.3%
18.0%
17.5%
14.2%
15.2%
13.2%
13.9%
Source: ADB (various issues) UN (various issues) IMF (various issues) and author’s estimation.
Chapter 6 __________________________________________________________________Page 152
Table A6.1.2: simulation results of the aid-growth nexus model: 19801997
Year
1980
1981
1982
1983
1984
1985
1986
1987
1988
GY
1.2%
3.4%
2.2%
2.1%
2.3%
3.5%
4.9%
-1.1%
-1.8%
20.6%
IGY
9.1%
8.1%
16.2%
11.4%
9.9%
12.1%
9.4%
8.4%
IPY
1.1%
1.2%
1.2%
1.3%
2.0%
1.8%
2.7%
3.9%
4.2%
XY
7.0%
5.2%
8.3%
8.1%
8.3%
8.9%
8.0%
7.5%
10.4%
MKY
7.9%
8.6%
12.4%
10.8%
8.5%
10.2%
8.4%
5.5%
7.7%
IY
10.2%
9.4%
17.4%
12.7%
11.9%
13.9%
12.2%
12.3%
24.8%
MY
22.9%
24.6%
27.5%
29.7%
30.6%
32.2%
27.1%
25.3%
26.7%
CAY
-15.9%
-19.4%
-19.2%
-21.6%
-22.3%
-23.3%
-19.1%
-17.7%
-16.4%
GYS
2.3%
4.0%
3.0%
5.0%
2.6%
5.1%
3.9%
-2.2%
-1.4%
IGYS
8.7%
8.5%
12.6%
9.7%
11.2%
12.2%
11.9%
11.3%
17.0%
IPYS
1.6%
2.6%
2.1%
2.0%
3.1%
1.8%
2.5%
2.9%
3.6%
XYS
1.9%
5.8%
7.5%
7.5%
10.1%
7.7%
6.3%
8.1%
10.6%
MKYS
7.4%
9.6%
10.1%
10.4%
9.8%
8.1%
8.2%
7.3%
7.3%
IYS
10.3%
11.1%
14.7%
11.7%
14.3%
14.0%
14.4%
14.2%
20.6%
MYS
22.3%
25.6%
25.3%
29.3%
31.8%
30.1%
26.9%
27.0%
26.4%
CAYS
-20.5%
-19.8%
-17.8%
-21.7%
-21.7%
-22.4%
-20.6%
-18.9%
-15.8%
Year
1989
1990
1991
1992
1993
1994
1995
1996
1997
GY
13.4%
6.7%
4.0%
7.0%
5.9%
8.2%
7.0%
6.9%
6.9%
IGY
15.4%
12.0%
9.6%
9.7%
10.0%
11.6%
11.5%
11.5%
11.3%
IPY
3.8%
3.2%
5.3%
7.6%
12.1%
14.0%
13.7%
19.1%
17.4%
XY
8.9%
9.2%
10.3%
11.2%
18.2%
19.5%
17.6%
17.4%
15.8%
MKY
11.3%
10.2%
13.2%
13.6%
16.1%
16.9%
15.0%
17.5%
18.6%
IY
19.2%
15.2%
14.9%
17.4%
22.1%
25.6%
25.2%
30.6%
28.7%
MY
27.1%
21.6%
18.1%
22.8%
32.6%
36.6%
33.4%
37.2%
32.3%
CAY
-18.2%
-12.4%
-7.8%
-11.6%
-14.4%
-17.1%
-15.8%
-19.8%
-16.5%
GYS
7.5%
6.5%
3.1%
5.3%
4.0%
8.6%
6.1%
9.8%
7.0%
IGYS
17.8%
9.6%
9.3%
9.0%
12.2%
8.4%
14.4%
10.9%
11.3%
IPYS
3.7%
4.2%
5.4%
6.3%
11.1%
13.0%
14.7%
19.4%
17.4%
XYS
8.2%
9.7%
10.9%
12.8%
15.4%
17.5%
19.5%
19.4%
15.0%
MKYS
10.5%
11.1%
12.7%
13.7%
15.8%
15.8%
17.4%
17.7%
17.4%
IYS
21.5%
13.8%
14.7%
15.3%
23.2%
21.4%
29.1%
30.3%
28.7%
MYS
26.4%
22.5%
17.6%
22.9%
32.3%
35.5%
35.8%
37.4%
31.1%
CAYS
-18.2%
-12.8%
-6.7%
-10.1%
-16.9%
-17.9%
-16.3%
-18.0%
-16.2%
Legend: the letter S attached to the following variables stands for their simulated values, GY is output growth, the letter
Y is attached to the following variable to indicate the measurement of that variable in percent of GDP: IGY is public
investment, IPY is private investment, XY is total exports, .MKY is imports of capitals, IY is total investment, MY is total
imports, CAY is current account balance.
Chapter 7 __________________________________________________________________Page 153
Chapter 7
THE EFFECT OF UNSTABLE AID FLOW
ON ECONOMIC GROWTH
Although significant policy changes have been made in many
LDCs, the new policy environment does not deliver sufficiently
high growth rates to make significant inroads into poverty
except where the external trade environment is favourable and
reforms are adequately or stably financed.
UNCTAD (2000a, p. 127).
Institutions affect economic performance by determining
(together with the technology employed) the cost of transacting
and producing. They are composed of formal rules, of informal
constraints and of their enforcement characteristics; while
formal rules can be changed overnight by the polity, informal
constraints change very slowly.
North (1997, p. 1).
7.1 Introduction
The analysis carried out in Chapters 5 and 6 reveals that a high level and
stable aid flow contributes to economic growth. The preliminary success in
implementing the “adjustment programmes” has also led to the Lao
Government increasingly receiving financial aid. Higher aid inflows have
enabled the Lao Government to invest more than their savings. Therefore, it
is not surprising that the Lao PDR has achieved a higher growth rate of real
income during the 1990-97 period. However, as pointed out in Chapter 1, the
decline in the growth rate of real income in the Lao PDR after 1997 has been
attributed to the negative impact of the Asian financial crisis and the
instability of aid and FDI inflows. This unfavourable effect on economic
growth has raised concern as to whether the economic development
achieved in the Lao PDR can be sustainable.
Chapter 7 __________________________________________________________________Page 154
As explained in Chapter 3, the weakness of policy conditionality and the lack
of state and institutional capability of the recipient country are the main
factors that cause economic volatility and unstable aid inflow. This
unfavourable environment in turn caused the LDCs to fail to achieve
economic growth with external viability. This chapter applies this explanation
to examine the effect of unstable aid flows on economic growth in the case of
the Lao PDR. The linkage between various types of political regimes and
economic growth is also explained here. By doing this, some light could be
shed on how adopting and implementing the policy and institutional reforms
may affect the Lao Government’s ability to maintain economic growth with
external viability. The macroeconometric model employed in Chapter 6 is
modified to use for the counterfactual analysis. The rest of this chapter is
organised as follows. Section 7.2 briefly explains the role that states and
institutions play in the process of economic growth. Section 7.3 examines
key institutional issues and the influence on the Lao Government’s ability to
raise
investment
with
external
viability.
Section
7.4
applies
the
counterfactual technique to disentangle the potential effects of unstable aid
flow from the effect of the Asian financial crisis on economic growth. The
chapter ends with some concluding remarks in Section 7.5.
7.2 States, institutions and economic growth
In an effort to become creditworthy, developing countries have altered their
economic policies and moved decisively in the direction of economic
liberalisation. Countries entering the “adjustment programmes” have done
quite well in terms of policy reform, in that the LDCs “have kept up with the
reform of public enterprise. In the areas of pricing and marketing reform, the
LDCs have gone further than the other developing countries” (UNCTAD,
2000a, p. 103). However, according to the World Development Report 1997
(World Bank, 1997) many governments in developing countries have failed
“to ensure law and order, protect property, and apply rules and policies
predictably” (ibid., p. 5). Many governments in developing countries also
found “it difficult to achieve the strong fiscal and monetary discipline required
Chapter 7 __________________________________________________________________Page 155
for economic stability” (ibid., p. 46). Therefore, rather than trying to evaluate
whether or not the policy changes make a contribution to economic growth, it
is now more important to question whether such policy changes are
effectively enforced by the recipient country. With this question, it is also
useful to examine the problems of policy mismanagement posed by the lack
of state and institutional capability. These problems have attracted much
attention in the development literature recently. The interest in exploring the
observed unevenness in implementing policy reforms under the Washington
consensus strategy is mainly focused on the problems of the lack of state
and institutional capability (Wallis and Dollery, 2001, p. 1).
According to the World Development Report 1997, states and institutions are
a crucial factor in determining economic outcomes (World Bank, 1997).
Figure 7.1 below illustrates how states, institutions and economic outcomes
are related.
Figure 7.1: The state, institutions, and economic outcomes
Culture, history
Informal rules
and norms
Institutional
structure
Formal rules
Legislativ e

State Executive
Judicial

Behaviour of
State agencies
Incentive structure
(including property
rights)
Transactions
costs
Technology
Contracts
Economic outcomes
Sources: World Development Report (World Bank, 1997, Figure 2.1, p.30)
Chapter 7 __________________________________________________________________Page 156
North (1990, p. 4) defines institutions as “any form of constraint that human
beings devise to shape human interaction”. Essentially, institutional
structures can be determined by both formal and informal rules. Formal rules
are laws and regulations governing the interaction between individuals,
businesses and organisations in societies including ministries, city councils,
universities, trade unions, churches, firms, foreign investors and the like.
Examples of formal rules include laws covering constitutions, property rights,
contracts, budgets, taxation, foreign investment, regulations governing the
operation of the central bank and the like. Informal rules are not legally
codified; they include socio-religious and cultural norms and values.
Examples of informal rules include norms and values related to honesty,
hard work, respect for chiefs and the like (ibid.).
Institutions are designed to facilitate human beings to produce and to
exchange goods and services within societies. According to North (1997, p.
6) “a basic function of institutions is to provide stability and continuity by
dampening the effects of relative price changes. It is institutional stability
that makes possible complex exchange across space and time”. Other
functions of institutions as specified by the World Bank (2001, p. 8) are as
follows:
(i)
They channel information about market conditions, goods and
participants. A country that provides good information flows helps
businesses to identity partners and find high return activities. The
information flows also help governments to regulate well.
(ii)
They define and enforce property rights and contracts, determining
who gets what and when. With property right assignments and the
ability to protect these rights, institutions can reduce the potential
for disputes and help enforce contracts.
(iii)
They help to increase or decrease competition in markets. This
function gives people and businesses incentives to do better and
promote equal opportunities.
Chapter 7 __________________________________________________________________Page 157
The state or the government is normally defined in three distinct sets of
powers. “One is the legislature, whose role is to make the law. The second is
the executive (sometimes referred to as government), which is responsible
for implementing the law. The third is the judiciary, which is responsible for
interpreting and applying the law” (ibid., 1997, p. 20). As can be seen from
Figure 7.1, state agencies together with the institutions and technology
employed determine the economic outcomes though their interactive effects
on transaction costs and contracts. In this respect, a country needs both
effective institutions and capable state agencies to ensure that sound
economic management can be promoted and that sustainable economic
growth can be achieved (i.e., good economic outcomes).
However, the conundrum remains as to which political regime can promote
such
abovementioned
economic
outcomes.
The
past
development
experiences of developing countries have revealed that a democratic or
authoritarian regime cannot reasonably claim to have any comparative
advantage when it comes to effective institutions and capable state
agencies. Hutchcroft (1994) demonstrates that one can find weak states in
both democratic and authoritarian regimes.1 Nevertheless, the past
development experiences of developing countries have also revealed that
cultural and historical differences across countries resulted in a diversity of
institutions and state regimes. Therefore, there can be many institutional and
political alternatives for managing the development process soundly. The
type of political regimes declaimed the best for promoting sound economic
management often varies and depends on the values and cultural
backgrounds of the declaimer. As Weder points out:
Somebody with Anglo-Saxon background would probably think
in terms of constitutions, and checks and balances. The Swiss
will be tempted to think that direct democracy and federalism
would be the solution for unstable countries. A follower of the
1
Siermann (1998) finds no causal relationship exists between political freedom and economic growth in a sample of 121
countries. On the other hand, Nelson and Singh (1998) cross-sectional study of 167 countries notes that a lack of
democracy and political freedom seriously hurts a nation’s economic performance. Gounder’s (2002) time-series study
of Fiji points out that Fiji’s non-democratic political environment and uncertain economic policies severely damaged its
growth prospects.
Chapter 7 __________________________________________________________________Page 158
Asian way would probably plead for strong bureaucracies and
corporatist structures.
Weder (1998, p. 354)
While there remains controversy about the linkage between different types
of political regimes and their influence on sustainable economic growth, the
emerging consensus on this issue is that state agencies have an important
role to play in promoting good governance in order to maintain sustainable
economic development.2 According to the World Bank (1997, p. 5) “the state
is essential for putting in place the appropriate institutional foundation for
market”. For institutions to be effective and meaningful it requires a capable
state to set the rules that facilitate collective actions between individuals,
businesses and organisations within the society, and to ensure that the rule
of law is strictly and consistently enforced (ibid., p. 34).
The interlinkage between states, institutions and the elements of good
governance tends to be mutually supportive and reinforcing. State capability
cannot be assured in the absence of accountability. As a lack of
accountability tends to reduce the credibility of states, this makes it hard to
predict the government’s decision making on policy changes and the way it
will respond to external shocks. Participation can be related to accountability
to some extent. For instance, if citizens are allowed to select their
representative through the electoral process, this participation principle will
encourage public officials to be accountable to the electorate, making states
more accountable in turn. Transparency cannot be assured in the absence
of an effective institutional framework that balance the right of disclosure
2
In recent years, the promotion of good governance is high on the policy dialogue agenda between donors and
recipients. According to ADB (1995, pp.8-12), four basic elements of good governance are defined as follows: (i)
Accountability is imperative to make public officials answerable for government behaviour and responsive to the entity
from which they derive their authority. It also means establishing criteria to measure the performance of public officials,
as well as oversight mechanisms to ensure that the standards are met. (ii) Participation is a principle to foster a “bottomup” approach to economic and social development. In this context, individuals, various associations and organisations
are allowed to give voice through letters to newspaper editors, participating in radio and television talkshows, and voting
in order to improve the design and implementation of public programmes and projects. (iii) Predictability refers to the
existence of laws, regulations and policies to regulate society. It also refers to their fair and consistent application. It
requires the state and its subsidiary agencies to be as much bound by, and answerable to, the legal system as are
private individuals and enterprises. (iv) Transparency refers to the availability of information to the general public and
clarity about government rules, regulations and decisions. Transparency in government decision-making and public
policy implementation reduces uncertainly and can help inhibit corruption among public officials.
Chapter 7 __________________________________________________________________Page 159
against the right of confidentiality. Predictability in the functioning of the
effective institutional framework could be useful to ensure the accountability
of the public sector. Predictability also requires ensuring that civil societies
adhere to the rule of equality before the law.3
Given the pivotal role of states, institutions and their interlinkage among the
four basic elements of good governance, it is hardly surprising that the lack
of state and institutional capability to enforce the rule of law has retarded
economic growth in many developing countries. The World Development
Report 1997 remarks on this viewpoint as follows:
…States that change the rules frequently and unpredictably,
announce changes but fail to implement them, or enforce rules
arbitrarily will lack credibility, and markets will suffer
accordingly.
World Bank (1997, p. 34)
According to the World Development Report 1997, the lack of predictability,
which tends to reduce the credibility of states over time, affects not only the
business environment but also the environment for the implementation of
development projects. The lack of credibility reduces investment, growth and
the return on development projects as a result of which “entrepreneurs
choose not to commit resources in highly uncertain and volatile
environments, especially if those resources will be difficult to recover should
the business environment turn unfavourable” (ibid., 1997, p. 36).
Similarly, without effective political institutions to run countries it tends to
create incentives for corruption. The institutional failure in this regard
hampers the effective delivery of public services, impedes the efficient use of
government revenues, discourages private investment and hence limits
economic growth (Mauro, 1997). In addition, as corruption is the antithesis of
accountability and transparency, the World Bank (2001) indicates that
political institutions that advocate free and fair elections tend to reduce
3
The issues of governance and sound developement have been discussed in detail by the ADB (1995).
Chapter 7 __________________________________________________________________Page 160
corruption. In democracies, citizens who are faced with cronyism and corrupt
politicians can influence the individual politician’s accountability through the
ballot box. In contrast, political institutions that restrain press freedom tend
to create an opportunity for misconduct in the public sector. Without detailed
information on government services, citizens are vulnerable to bureaucratic
harassment and public officials may arbitrarily act for their own benefit in
terms of overcharging fees or demanding bribes.4
Turning to the problems of policy mismanagement posed by the lack of
states and institutional capability, the World Development Report 1997
remarks that “although the recipe for good policies is well known, too many
countries still fail to take it to heart and poor performance persists. This often
signals the presence of political and institutional incentives for maintaining
bad policies” (World Bank, 1997, p. 49). In countries with some form of
democracy, the mismanagement of macroeconomic policy could take place
in the presence of inefficient political cycles or political instability. The
public-choice school claims that democratic decision-making influences the
outcomes of budget deficits (Siermann, 1998).
According to the “opportunistic theory” (Nordhaus, 1975), as well as the
“partisan theory” (Hibbs, 1977), politicians use their monetary and fiscal
powers to manipulate the economy to their own benefit or to help their
constituency. Right wing governments, despite a stated preference for low
levels of government spending, tend to run budget deficits by increasing
government spending in order to increase the probability of staying in office
(Persson and Svensson, 1989). Left wing governments tend to undertake
ambitious programmes and give priority to distributive objectives. This
political behaviour is likely to lead to the problem of macroeconomic
populism which result in large government deficits, high inflation and an
unsustainable external imbalance (Hossain and Chowdhury, 1998). In
countries with coalition governments, Siermann (1998) argues that internal
4
See World Bank (2001, pp. 108-109) for a detailed discussion of this issue.
Chapter 7 __________________________________________________________________Page 161
policy conflicts are likely to take place more frequently. Therefore, “the
greater the internal political conflicts, the more difficult it will be to enact
deficit-reduction measures” (ibid., p. 2). Using a sample of 123 countries, Shi
and Svensson (2000) show that budget deficits increase by an average rate
of about 1 percent of GDP in the election year and that these deficits persist
for several years after the election. Using a sample of 60 countries, Persson
and Tabellini (2000) find that countries with coalition governments have
budget deficits about 1.5 to 2 times larger than other types of government.
It is now widely recognised that persistent large budget deficits in developing
countries are the root cause of external imbalances (Hossain and
Chowdhury, 1998). However, whether the external imbalance gives rise to a
balance-of-payments crisis is not straightforward. This effect depends on
how budget deficits are financed, the country’s savings level, the
effectiveness of financial institutions and the government’s ability to use
fiscal measures to handle the shocks that affect their budgets.5 The
combination of policy conditionality aiming to reduce domestic borrowing and
the extremely low rate of domestic savings in LDCs makes the LDCs heavily
reliant on foreign borrowing to finance budget deficits. Therefore, persistent
large budget deficits could lead to balance-of-payments crises, as “over
reliance on foreign borrowing can cause appreciating real exchange rates,
widening current account deficits, unsustainable external indebtedness and
dwindling foreign exchange reserves” (ibid., p. 132).
So far, the problems attributed to the way the lack of a country’s state and
institutional capability may affect investments and economic growth has
been drawn from the weakness of democratic political regime. Findings
indicate that a country’s state and institutional capability is a crucial factor in
determining economic development. A country that failed to ensure law and
order and failed to apply predictable policy tends to undermine its
creditability. A country that failed to prevent politicians from manipulating
5
See Hossain and Chowdhury (1998, pp. 146-152) for further detailed discussions on the linkage between budget
deficits and balance-of-payment crises.
Chapter 7 __________________________________________________________________Page 162
fiscal and monetary expansion tends to create persistent budget deficits.
Therefore, a country with a weak state and institutional capability tends to
create the aforementioned development failures, reducing investments and
also creating an unsustainable level in the balance of payments. From this
point of view, it can be said that the lack of state and institutional capability
can be a predominant factor impairing the developing countries’ ability to
maintain economic growth with external viability. The next section examines
the state and institutional problems under the Lao PDR’s authoritarian
political regime.
7.3 Foreign
aid,
the
state,
institutions
and
economic
performances in the Lao PDR
Unlike many other transition economies, the Lao PDR carried out economic
reform without abandoning a one-party political regime. This led to both
various opportunities and obstacles. To analyse this, Section 7.3.1 examines
the key governance issues focusing on the factors that undermined the state
and institutional capacity. Section 7.3.2 addresses the problems of the lack
of state capacity to deal with the trade-off between the need to boost
economic growth and the need to maintain external viability.
7.3.1 Key governance Issues in the Lao PDR
Since the establishment of the Lao PDR in 1975, the Lao People’s
Revolutionary Party (LPRP) has functioned without a written constitution. In
the absence of fixed and knowable law, this situation has added to the
reluctance of international development agencies to invest in the country.
This led to the Supreme People’s Assembly to adopt a constitution in 1991
(Savada, 1994, p. 219). Although the newly adopted constitution outlines a
system composed of the executive, legislative and the judicial branches, in
practice the LPRP continued to influence the governance and the choice of
leaders through its constitutional "leading role" at all levels (Savada, 1994,
Chapter 7 __________________________________________________________________Page 163
p. 224). Accordingly, the Party Congresses have been an important forum for
policy making and consensus building.6 The processes of policy formulation
are made through lengthy periods of internal debate, especially at the upper
echelons of the party. In this regard, the policy consensus building has
facilitated the implementation of the “adjustment programmes” without
causing political instability, and helped the Lao Government to quickly
stabilise the economy and moved the country in the direction of a liberal
economic regime.
Although the Lao Government had made progressive changes in the process
of the “adjustment programmes”, the enforcement of those policy changes
was held back by its lack of state and institutional capacity (World Bank,
1996). Attempts by the Lao Government to translate policy conditionality into
administrative actions faced the lack of legal systems and regulation
frameworks necessary to promote good governance. Wescott (2001, p. 26)
points out that these constraints exist at the individual, institutional and
organisational level. At the individual level, low wages do not motivate civil
servants to perform their jobs effectively. According to Wescott, this problem
is exacerbated by poor incentives in recruitment and promotion policies.
Also, recruitment and promotion tend to be based on clan ties, seniority and
political patronage rather than merit, talented staff tends to be marginalised
through transfers and assignment to low priority tasks. He further notes that
at the institutional level, there are no proper job descriptions, no clarification
of roles and responsibilities and no grievance resolution mechanisms.
Without these regulations, bureaucrats tend not to take risks and pass even
minor decisions to their superiors (ibid.). At the organisational level, Wescott
points out that the lack of co-ordination in public administration is due to the
6
As in other one-party states, the Party Congress generally provides good indications for the direction of policy and
institutional changes. The Fourth Party Congress in 1986 endorsed the New Economic Mechanism that aimed to
transform the centrally planned economic system to the market economy. The Fifth Party Congress in 1991 consolidated
the political power of the advocates of economic reforms. Institutions were established to secure the maintenance of the
one-party state. The Sixth Party Congress in 1996 reaffirmed the commitment to economic reform, but preferred a
slower reform path (Bourdet, 2000, pp.2-3). The Seventh Party Congress in 2001 set the development strategy for
getting the country out of the state of underdevelopment by the year 2020. The congress also emphasised the need to
apply the systematic state-managed market economy in order to gradually shift a fundamental economic structure
towards an industrialised and modern economy (Lao People’s Revolutionary Party, 2001).
Chapter 7 __________________________________________________________________Page 164
application of the bottom-up reporting system without institutionalised
channels for interministerial communication. In addition, silence about the
party’s functions and powers in the constitution runs parallel to the
legislative, executive and judiciary branches that limit the independence of
these three powers from each other.
As indicated in the preceding section, the interlinkage among the state,
institutions and the four basic elements of good governance are mutually
supportive and reinforcing. However, it is important to indicate that the
capacity constraint at the institutional level may undermine the Lao
Government’s ability to create a governance system with accountability,
transparency and predictability. Indeed, Wescott (2001) regards the system
of administration in the Lao PDR as an opaque regime where “nepotism and
patronage are endemic within the party and the bureaucratic system. The
lack of transparency makes it extremely difficult for outside observers to fully
understand the political decision-making structure” (ibid., p. 25). In addition,
the lack of accountability and transparency creates rampant corruption at all
levels of officialdom. “This behaviour reduces government revenues,
misallocates expenditure, reduces foreign investment and erodes public
trust” (ibid. 30).
Similarly, it is arguable that the capacity constraints at the individual and
organisational levels together with the domination of the LPRP could
undermine the Lao Government’s administrative capacity to deliver its
functions. For example, “the 1990 Central Bank Law gave the Central Bank
formal autonomy over monetary and exchange rate policy, [however] the
bank works under strict political control” (Wescott, 2001, p. 28). The
domination of the LPRP in the Central Bank’s decision-making process
coupled with the lack of co-ordination within interministerial administration
“made the impact of the regional economic downturn on the macroeconomic
situation in the Lao PDR considerably worse than it could have been”
(Bourdet, 2001, p. 166).
Chapter 7 __________________________________________________________________Page 165
7.3.2 Foreign aid and the Lao Government’s ability to raise investment
with external viability
The above discussions indicate that the monopoly power of the Lao People’s
Revolutionary Party has significantly influenced the pace and progress of the
“adjustment programmes”. This opportunity has facilitated the elaboration,
adoption and implementation of economic reform. However, the failure to
regulate economic and political interactions posed by the lack of state and
institutional capacity has undermined the Lao Government’s ability to
maintain sustainable economic growth. These issues are further examined in
this section in order to have a clear understanding of how the
abovementioned opportunities and obstacles may affect the Lao PDR’s
ability to maintain economic growth with external viability.
To start with, the opportunities created by the Lao Government during the
implementation of economic reforms, after entering the “adjustment
programmes” in 1989, moved the country decisively in the direction of
economic liberalisation. The Lao Government’s commitment to reducing
budget deficits and tightening monetary policy at the outset of the
“adjustment programmes” quickly brought stability to the macroeconomic
environment. As can be seen from Table 7.1 below, the ratio of budget
deficits to GDP reduced from its peak level of an annual average of about 15
percent during the 1989-90 period to 10 percent during the 1990-96 period.
Money growth has also fallen from its peak level during the 1986-88 period
at an annual average of about 86 percent to about 49 percent during the
1989-90 period and remained low at about 34 percent for the 1990-96
period. Inflation was gradually brought down from an annual average of
about 50 percent during the 1989-90 period to 12 percent during the 199196 period. The exchange rate stabilised, with depreciation of the exchange
rate declining from an annual average rate of about 50 percent during the
1989-90 period to 5 percent during the 1991-96 period.
Chapter 7 __________________________________________________________________Page 166
Table 7.1: Selected macroeconomic indicators 1978-96
Year
Money Growth
Public Investment Growth
Budget Deficits
Foreign Aid/GDP
FDI/GDP
Foreign Currency Reserve*
Inflation
Exchange rate Depreciation
Real GDP growth
1978-85
36%
27%
-13%
12%
0%
73%
66%
2%
1986-88
62%
65%
-11%
9%
0%
1.7
39%
50%
1%
1989-90
49%
-16%
-15%
19%
1%
3.7
50%
33%
10%
1991-96
34%
6%
-10%
16%
4%
3.3
12%
5%
6%
Note: * equivalent months of imports
Source: ADB (various issues) and IMF (various issues)
As the disbursement of aid is subject to the performance of conditions in the
“adjustment programmes”, success in implementing the stabilisation
programmes in the early 1990s led to a surge of aid inflow. The ratio of aid
to GDP almost doubled from the pre-“adjustment programmes” level to
remain at about 16 percent for the 1991-96 period. The other sources of
capital inflow such as foreign direct investment (FDI) were also increased, as
foreign direct investment has been promoted since 1988. The ratio of FDI to
GDP increased sharply from 0.4 percent in 1988 and reached its peak at 8.6
percent in 1996. Consequently, the growth rate of real GDP increased
fivefold from an annual average rate of about 2 percent during 1978-85
period to 10 percent for the 1989-90 period. The real economic growth rate
remained strong at an annual average rate of about 6 percent for the 199196 period.
It should be noted that economic growth during the 1991-96 period was also
accompanied by expansionary fiscal and monetary policy to satisfy an
increase in the demand for domestic investment. An annual average rate of
money growth was about 34 percent. The growth rate of public investment
also increased from a negative annual rate of about 16 percent in the 198990 period to a positive annual rate of 6 percent for the 1991-96 period. This
led to a rapid increase in the demand for both capital and consumption
goods. Accordingly, the current account deficit also gradually deteriorated
from –3.8 percent of GDP in 1992 to –11.9 percent to GDP in 1996. As
illustrated in Figure 7.2 below, the deterioration of the current account
Chapter 7 __________________________________________________________________Page 167
balance in this period reflects the fact that the growth in export earnings was
unable to keep pace with a rapid surge in imports of both capital and
consumption goods. As part of the external account deficit had been
financed by foreign aid and other sources of foreign capital inflows, the
country’s foreign exchange reserves and foreign exchange earnings must be
utilised to finance the remainder of the external account deficit in order to
maintain external viability. However, as demonstrated in Chapter 6, the Lao
PDR’s ability to raise foreign exchange earnings from exports is subject to its
access to markets and the demand for its export commodities. In fact, efforts
to raise foreign exchange earnings were often blocked by complex
international trade barriers and the volatile economic environments of the
Lao PDR’s trading partners. Therefore, export earnings are a crucial factor
dominating the Lao Government’s ability to raise investment with external
viability, as well as foreign exchanges from foreign aid and FDI inflow.
Figure 7.2: Trade and current account balance, 1989-2001
Total exports as % of Total imports
2001
2000
1999
1998
1997
1996
1995
1994
1993
1992
1991
1990
1989
70.0%
60.0%
50.0%
40.0%
30.0%
20.0%
10.0%
0.0%
-10.0%
-20.0%
-30.0%
Current account balance as % of GDP
Source: Data extracted from Appendix 7.1, Table A7.1.1
The need to promote economic growth by running an expansionary fiscal
and monetary policy often creates the deterioration of external balance in
developing countries (Hossain and Chowdhury, 1998). To maintain external
viability, the government needs to borrow money from the IMF to finance the
deficit. Such borrowing allows the government more time to undertake the
necessary adjustment. Under the IMF’s policy prescription, a country needs
to devalue its nominal exchange rate along with implementing prudent fiscal
Chapter 7 __________________________________________________________________Page 168
and monetary policy to ensure that the external current account balance is
viable. However, the obstacles facing the Lao PDR’s efforts in implementing
policies to adjust the external balance are rooted in the lack of state and
institutional capabilities. These issues are discussed below.
From 1996, the central bank has periodically attempted to slow down money
growth but it failed to prevent the growing level of credits extended to the
private sector and state enterprises. This failure was a consequence of the
lack of institutional capability in the financial system of the Lao PDR, as the
banking system “remains underdeveloped and weak”, and “the banking
supervision framework has not been able to correct major bank weakness in
a timely manner” (IMF, 1998b, p. 2). In the wake of the Asian financial crisis
in 1997, the fluctuations in tax collection mainly caused by the weakness of
fiscal institutions and a chronic lack of transparency in the state and
businesses made it hard for the government to meet expenditure targets
(i.e., counterpart funds required to implement projects financed by aid
agencies). Despite this, the government did not adjust its expenditure plans
but increased investment in an irrigation project in order to promote selfsufficiency of rice production. In addition, the decline in export earnings
since 1995 led to a deterioration in the current account balance. This
situation became worse during the Asian financial crisis, as the Lao PDR
faced unsustainable balance-of-payment positions. The IMF’s decision not to
provide financial support during the crisis was because there were no
contingency measures for unexpected events. The IMF also disagreed with
the irrigation investment plan, as the expansionary fiscal policy involved
could worsen the external account imbalance.7 Other aid donors also cut
their financial aid because the pace of economic reform in the Lao PDR
slowed down.
7
The “adjustment programmes” is designed to support external account deficits at the level that could be financed by
normal capital inflow. Expansionary fiscal policy in the wake of the Asian financial crisis conflicted with this objective.
Therefore, it was not suprising that the IMF did not provide the financial support.
Chapter 7 __________________________________________________________________Page 169
Given financial shortfalls in both internal and external revenues, the Lao
Government financed increased public spending by money creation and the
drawing down of foreign reserves. Expansionary fiscal and monetary policy
thus widened the budget deficit.
2001
2000
1999
1998
1997
1996
1995
1994
1993
1992
1991
1990
30.0%
120.0%
20.0%
100.0%
80.0%
10.0%
60.0%
0.0%
40.0%
-10.0%
20.0%
-20.0%
0.0%
Money growth
Budget deficit as % of GDP
Money Growth
Budget Def. & Public Inv.
Growth
1989
Figure 7.3: Fiscal and monetary policy and budget deficits, 1989-2001
Public investment growth
Source: Data extracted from Appendix 7.1, Table A7.1.1
As illustrated in Figure 7.3, the growth rate of money and public investment
reached their peaks at about 113 percent and 9 percent per annum,
respectively. The budget deficit increased from 5 percent of GDP in 1997 to
9 percent of GDP in 1998. Given the high degree of currency substitution in
the Lao economy where people preferred to hold the Thai Baht and the US
dollar to the local currency, expansionary fiscal and monetary policy had
multiplied the already volatile macroeconomic environment (Iweala, et al.,
1999). By the end of 1997, the Lao currency has lost about 183 percent of its
value against the US dollar, while inflation increased by 19.3 percent (see
Figure 7.4 below).
Chapter 7 __________________________________________________________________Page 170
Figure 7.4: Inflation and exchange rates variation 1985-2001
200.0%
150.0%
100.0%
50.0%
0.0%
Inflation
2001
2000
1999
1998
1997
1996
1995
1994
1993
1992
1991
1990
1989
-50.0%
Exchange rate depreciation
Source: Data extracted from Appendix 7.1, Table A7.1.1.
In an attempt to stabilise the macroeconomic environment without financial
supports from the IMF, the Lao Government encountered difficulties due to
the lack of legal and regulatory frameworks necessary for the effective
functioning of fiscal and monetary institutions. This exacerbated the “lengthy
consensus building in the decision-making process [at the upper echelons of
the party], making it difficult for the country’s authorities to react quickly to
the deteriorating macroeconomic situation” (Iweala, et al., 1999, p. 2).
Figure 7.5: Foreign capital inflows and current account balance 19852001
25.0%
20.0%
15.0%
10.0%
5.0%
Growth of real GDP
Aid as % of GDP
2001
2000
1999
1998
1997
1996
1995
1994
1993
1992
1991
1990
1989
0.0%
Foreign direct investment as % of GDP
Source: Data extracted from Appendix 7.1, Table A7.1.1.
Essentially, after the recovery of Thai Baht in 1998, the Lao Kip depreciated
60 percent in 1998 against the US dollar, while the inflation rate reached its
peak at 86.1 percent. Efforts to tighten fiscal and monetary policy had been
made in late 1998 and early 1999. This led to a decline in annual inflation to
Chapter 7 __________________________________________________________________Page 171
51.2 percent in 1999, while the Lao currency lost another 77.8 percent of its
value against the US dollar. The instability of aid inflows and the decline in
FDI since 1997 led to a decline in the annual growth rate of real GDP at 4
percent for the 1998-1999 period. The rate of economic growth after the
Asian financial crisis represents about half of the annual average growth rate
during the 1991-97 period. This unfavourable circumstance raised concerns
over sustainable development in the Lao PDR.
7.4 Counterfactual analysis: the effect of unstable aid inflow
on economic growth in the case of the Lao PDR
The preceding sections have demonstrated that the failure to ensure viable
balance of payments was not only caused by the weakness of the IMF’s
policy design but also by the Lao Government’s policy mismanagement
exacerbated by the lack of state and institutional capacity. These problems
led to the instability of foreign aid and FDI inflows and in turn worsened the
adverse effects of the Asian financial crisis. The following sections examine
the
effects
of
unstable
aid
inflows
on
economic
growth
in
the
macroeconometric framework. The counterfactual technique is employed to
disentangle the potential effects of unstable aid flow from the effects of Asian
financial crisis on economic growth. Section 7.4.1 explains the model
specification and the alternative assumptions of how to evaluate the effects
of unstable aid inflow on economic growth. The results are discussed in
Section 7.4.2.
7.4.1 The model and policy experiments
To assess the potential effects of unstable aid inflow on economic growth,
the supply side macroeconometric model employed in Chapter 6 is modified,
in which public investment is treated as policy variable (exogenous variable).
In this context, the model has four behavioural equations and three identity
equations. The specifications of the model in the generic form are as follows:
Chapter 7 __________________________________________________________________Page 172
A specification of the aid-growth nexus model
Behavioural Equations
Output Growth
GYt = f1 ( IYt ; GLEt )
(7.1)
Investment
IPYt = f 3 ( GYt , IGYt ; FDIYt , INFt )
(7.2)
International Trade
XYt = f 4 ( MKYt ; YFt REERt )
MKYt = f 5 ( IYt ; REERt )
(7.3)
(7.4)
Identity Equations
IYt = IGYt + IPYt
(7.5)
MYt = MKYt + MCYt
CAYt = XYt − MYt + U t
(7.6)
(7.7)
where GY:
IY:
IGY:
IPY:
XY:
MKY:
MCY:
MY:
CAY:
AY:
INF:
GLE:
YF:
growth rate of income,
the share of investment in GDP,
the share of public investment in GDP,
the share of private investment in GDP,
exports to GDP ratio,
the share of imports of capitals in GDP,
the share of imports of consumption goods in GDP,
the share of imports to GDP,
the ratio of current account deficit to GDP,
the ratio of total aid to GDP,
inflation rate,
the growth rate of effective labour force,
the weighted average by trade share of growth in per
capita income of the trading partners,
REER: real exchange rate, and
U:
the error term including net service transfers.
As mentioned earlier, the weakness of the IMF’s conditionality and the lack
of state and institutional capability of the Lao PDR are the main factors
leading to unstable aid and FDI inflows. To evaluate the adverse effects of
unstable aid flow on economic growth, it is necessary to create a baseline
Chapter 7 __________________________________________________________________Page 173
model and perform policy experiments. The baseline model is used to
generate an approximate actual outcome for all endogenous variables
during the 1998-2001 period. As such, it is required to employ actual data of
all exogenous variables for the same period to perform the ex-post
simulations. This approach has a minor problem due to the lack of
unemployment and student enrolment data for the 1997-2001 period to
compute the growth rate of in the effective labour force.8 This problem has
been solved by using the goal-seeking technique, i.e. the growth rate in the
effective labour force has been used to generate the approximation of the
historical trends for all endogenous variables. Therefore, the solution for the
growth rate of in the effective labour force will implicitly capture the side
effect of unstable aid and FDI inflows and the Asian financial crisis on output
growth and on current account deficit for the 1998-2001 period.9
Policy experiments are used to examine how the improvement of donors’
conditionality and the Lao PDR’s state and institutional capability may have
an impact on current account deficits and output growth. Thus, the
experiments are designed to simulate the effects of stable aid flows in order
to capture the adverse effects of unstable aid flow. They are carried out
under different assumptions as follows:
•
Experiment (e1) aims to answer the question, “What would have been the
outcome of output growth and external account deficits if the IMF had
bailed out the Lao Government from the Asian financial crisis?”10 This
experiment is carried out under the assumption that the IMF includes the
contingency measure in the “adjustment programmes”, which enables the
IMF to provide financial support for the balance-of-payments adjustment
to the Lao Government during the Asian financial crisis. Also assuming
8
As explained in Chapter 4, the growth rate of effective labour force was constructed by assuming that the rate of
unemployment is constant. This assumption does not hold for the 1998-2001 period, as it is expected that
unemployment will rise during the Asian financial crisis.
9
See Table A7.2.1 for detailed data employed in this simulation.
10
This experiment is based on a recent reconsideration to modify conditionality under the policy-based lending by the
International Monetary Fund (IMF) and World Bank from the “conditionality” to the “partnership” principle. This move has
been seen as an appropriate approach that could improve the effectiveness of aid. Unlike being under “conditionality”, the
proposal to carry out “partnership” principle lets the aid recipient take the lead in the process of policy formulation and in
the pace of policy implementation. It also allows for differences in perspectives and provides room for partners to learn
from mistakes. It is likely that the implementation of the “partnership” principle will reduce the tension over national
sovereignty issues and help to create a stable aid flow. See UNCTAD (2000a) for detailed discussions about the conflict
between donors’ conditionality and recipient country’s national sovereignty issues.
Chapter 7 __________________________________________________________________Page 174
that other donors provides financial support for irrigation projects to help
the Lao Government to achieve the promotion of self-sufficiency in rice
production. With this financial support (i.e. a stable aid inflow), not much
pressure would be put on the Lao Government to use inflation tax or to
draw foreign reserves to finance the budget deficits. Also, this financial
support would allow the Lao Government to adjust the exchange rate in
line with the Thai currency, which is freely circulated along with the US
dollar in the Lao economy.11 After the Asian financial crisis the Lao
Government can maintain a stable macroeconomic environment and
expand credit to satisfy investment demands from the indigenous
investors without facing difficulty in the balance-of-payments. As such the
experiment is carried out under the following assumptions: Inflation is set
to follow its actual trend for the 1997-1998 period, and grows at 12
percent annually for the 1999-2001 period. The real effective exchange
rate is set to follow its alignment (i.e., REER=100 for each year). Foreign
direct investment and public investment variables are set to follow their
historical paths.12 Finally, the growth rate in the effective labour and other
exogenous variables follow the same path as in the baseline
experiment.13
•
Experiment (e2) aims to answer the question, “What would have been the
outcome of output growth and the external account balances, if the Lao
Government had managed to raise investment with external viability?” As
explained in the preceding section, a country’s state and institutional
capability plays a crucial role in improving the country’s ability to raise
investments with external viability. It is arguable that strengthening the
institutional capacity could help the Lao PDR to better cope with external
shocks. Strengthening the institutional capacity could also increase the
Lao PDR’s credibility and make the Lao administration more effective.
This in turn could help the Lao PDR to attract foreign investment though
creating a favourable economic environment.14 Therefore, foreign direct
11
Note that after the Lao Government regained control over fiscal and monetary policy in the late 1999, it relied on a free
interest loan from China together with the promotion by the Lao Government to use local currency for bilateral trade with
China and Vietnam, to stabilise the Lao currency. Thus, inflation was gradually stabilised by the year 2000 which enabled
the Lao Government to reaccess financial support from the IMF in early 2001.
12
Basically, the Lao Government has adopted a five-year plan to implement the public investment programmes (PIP).
Thus, it is pausible to assume that the Lao Government might try to finance the public investment either from foreign aid
or other sources to achieve the five-year plan for PIP, as it was surely the case that the Lao Government financed the
public investment by printing money and drawing down foreign reserves during the 1997-98 period, when they did not
receive financial aid to finance the irrigation project. Then, it will be pausible to assume that the public investment will not
deter from its historical trend, if the Lao Government could have used financial aid to finance the irrigation project in order
to achieve the five-year plan for PIP.
13
14
See Table A7.2.2 for detailed data employed in this simulation.
There are volumes of economic literature that provide the explanation for factors determining the movement of FDI
inflow across countries. Among others, these factors include the neoclassical theory of capital mobility, the industrial
organization approach, the transaction cost or internationalisation theory, Dunning’s eclectic theory and Komima’s
macroeconomic theory of FDI (Sun, 1998). However, the assumption of factors determining the FDI inflows in this
experiment is based on the view that policy conditionality attached to aid could help the recipient country to attract foreign
private capital. Krueger (1998, p.2010 & p. 2007) remarks on this view that “the present [IMF] programme provides
important signals for private creditors” and that “private creditors are often unwilling to extend credit lines unless the
[World Bank] and the [IMF] have first signalled their acceptance of economic policies”.
Chapter 7 __________________________________________________________________Page 175
investment is assumed to increase from its historical trend by 5 percent.
Other variables follow the same path as in experiment (e1).15
7.4.2 Counterfactual simulation results
The simulation results are shown in Figures 7.6 and 7.7 below. Figure 7.6
shows that the modification of IMF’ conditionality and the strengthening of
the Lao PDR’s state and institutional capability leads to the improvement of
the current account balance. These outcomes are captured by CAYEX1curve and CAYEX2-curve, respectively. As explained in Section 6.3.2, aid
and FDI inflows have contributed to export growth through the investment
channel. Therefore, an increase in aid and FDI inflow could have enabled
the Lao Government to increase export earning and improve the current
account balance.
Figure 7.6: Simulation of the impact of various experiments on current
account deficits
1995
1996
1997
1998
1999
2000
2001
-14%
-15%
-16%
-17%
-18%
-19%
CAYB
CAYEX1
CAYEX2
Source: Data extracted from Appendix 7.1, Table A7.1.2.
Figure 7.7 below also illustrates that the modification of the donors’
conditionality and the strengthening of the Lao PDR’s state and institutional
capability make a contribution to output growth. In experiment (e1), the
GYEX1-curve captures the benefits of the IMF’s financial bailout in the form
of a stable aid and FDI inflows, and also in the form of the stability of
15
See Table A7.2.3 for detailed data employed in this simulation.
Chapter 7 __________________________________________________________________Page 176
macroeconomic
environment.
With
this
financial
support,
the
Lao
Government would have been able to boost domestic investment. Thus, the
economy should have been better off than in the situation without the IMF’s
financial support. In experiment (e2), the GYEX2-curve captures the benefits
of strengthening the state and institutional capacity that could have helped
the Lao Government to attract more FDI inflow. This favourable
circumstance would have helped the Lao PDR to achieve a rapid growth of
real income.
Figure 7.7: Simulation of the impacts of various experiments on output
growth
10%
9%
8%
7%
6%
5%
4%
3%
1995
1996
1997
GYB
1998
GYEX1
1999
2000
2001
GYEX2
Source: Data extracted from Appendix 7.1, Table A7.1.2.
It is obvious from the counterfactual analysis that the IMF’s conditionality
and the Lao PDR’s state and institutional capability are important for
promoting sustainable economic growth. However, as mentioned earlier, the
decline in real income and the difficulties to maintain external account
balance after the Asian financial crisis were attributed to the weakness of the
IMF’s policy design and the lack of state and institutional capability in the
Lao PDR. In fact, the approximations of actual outcomes of the current
account balance and output growth are respectively represented by the
CAYB-curve in Figure 7.6 and by the GYB-curve in Figure 7.7. As such, the
adverse effect of unstable aid flow on output growth can be measured by
comparing the simulation results of the baseline model with the simulation
Chapter 7 __________________________________________________________________Page 177
results of policy experiments. The gap between the GYB-curve and the
GYEX1-curve captures the adverse effect of unstable aid flow on output
growth arising from the weakness of donors’ conditionality. The gap between
the GYB-curve and the GYEX2-curve captures the adverse effect of unstable
aid flows on output growth arising from the weakness of the Lao PDR’s state
and institutional capability. The same measure can be applied for current
account deficits. As such, the gaps between the CAYB-curve and the
CAYEX1-curve and between the CAYB-curve and the CAYEX2-curve
capture the adverse effect of unstable aid flows on the current account
deficits.
From the above analysis it can be said that the negative effects of unstable
aid inflow and a volatile macroeconomic environment have impaired the
favourable impact of aid on economic growth. This adverse effect took place
due to lack of contingency measures in the IMF’s policy designs. This
unfavourable situation triggered the instability of aid flow, which was
intensified by the Lao PDR’s policy mismanagement posed by the lack of
state and institutional capacity. Therefore, the instability of aid and FDI
inflow could continue to affect the Lao PDR to achieve sustainable economic
growth.
7.5 Conclusion
This chapter has analysed the effect of how unstable aid inflow may affect
the sustainability of economic growth for the case of the Lao PDR. The
weakness of policy conditions designed for aid delivery and also the
problems posed by the lack of state and institutional capability for policy
implementation have been applied to explain the effectiveness of aid. The
analysis carried out here suggests that many symptoms of development
failures such as poor quality public services, rampant corruption, fiscal and
monetary indiscipline, uncontained current account deficits and the failure to
apply the rules and policies predictably can be attributed to the failure to
build good governance. This failure decreased the authority and legitimacy
Chapter 7 __________________________________________________________________Page 178
of the government and increased conflict over “rules of the game” for
economic and political interactions. These problems in turn have resulted in
many developing countries failing to maintain economic growth with external
viability. It has occurred regardless of the political regime. There is no
exception for the Lao PDR where a one-party political regime has the
“leading role” in policy formulation. Although the Lao PDR’s governance
system has ensured political stability during the implementation of the
“adjustment programmes”, the lack of a legal system and regulatory
frameworks remains at individual, institutional and organisational levels.
These constraints have retarded the Lao Government’s attempts to translate
policy conditionality into administrative actions. The silence of the LPRP’s
functions and powers in the constitution, coupled with the LPRP’s
interference into the legislative, executive and judiciary branches has also
damaged the Lao Government’s administrative capacity to deliver its
functions, and helped to trigger the instability of aid and FDI inflows. The
Asian financial crisis when combined with the aforementioned problems had
a drastic adverse effect on economic growth.
The effects of the problems caused by the weakness of policy conditionality
and the lack of state and institutional capacity of the Lao PDR have been
examined using the counterfactual analysis framework. The simulation
results show that the failure to improve the policy conditions designed for aid
delivery and also the problems of policy mismanagement posed by the Lao
PDR’s lack of state and institutional capability, have impaired the country’s
ability to raise investment with external viability. This failure in turn impedes
the Lao PDR’s efforts to maintain sustainable economic growth. These
findings also suggest that the instability of aid inflow tends to create
macroeconomic
instability
at
the
same
time
that
the
“adjustment
programmes” was designed to reduce it. The controversy between the
policies designed for aid delivery and the “adjustment programmes” played a
significant part in reducing aid effectiveness and in disrupting the
development process in the Lao PDR. These findings suggest that allowing
the recipient country to take the lead in the process of policy formulation and
Chapter 7 __________________________________________________________________Page 179
the pace of policy implementation, and also urging the recipient country to
strengthen their state and institutional capability would have reduced the
instability of aid flows and improved the effectiveness of aid.
Chapter 7 __________________________________________________________________Page 180
Appendix 7.1: Selected macroeconomic indicators and simulation
results of counterfactual analysis
This appendix consists of two tables. Table A7.1.1 presents some
macroeconomic indicators. Table A7.1.2 presents the results of policy
experiments to disentangle the negative effect of adverse effects of unstable
aid inflow from the adverse effect of the Asian financial crisis on economic
growth.
Table A7.1.1: Selected macroeconomic indicators: the Lao PDR, 19892001
Year
1989
1990
1991
1992
1993
1994
1995
Total exports as % of Total imports
32.7%
42.4%
56.8%
49.1%
55.7%
53.3%
52.8%
Current account balance as % of GDP
-16.6%
-9.7%
2.1%
-3.8%
-3.0%
-6.4%
-7.1%
Budget deficit as % of GDP
-16.4%
-13.5%
-10.6%
-10.0%
-9.2%
-10.8%
-10.1%
Public investment growth
-15.0%
-16.8%
-17.1%
8.8%
9.1%
25.5%
5.5%
Money growth
89.7%
7.5%
15.8%
49.0%
64.6%
31.9%
16.4%
Inflation
66.4%
33.1%
13.3%
9.3%
6.4%
7.7%
20.6%
Exchange rate depreciation
47.4%
18.4%
7.5%
-7.1%
0.7%
0.0%
13.0%
Growth of real GDP
13.4%
6.7%
4.0%
7.0%
5.9%
8.2%
7.0%
Aid as % of GDP
Foreign direct investment as % of GDP
20.3%
0.6%
17.6%
0.7%
15.2%
0.9%
13.9%
0.8%
15.6%
5.0%
14.1%
3.9%
17.3%
5.4%
Year
1996
1997
1998
1999
2000
2001
Total exports as % of Total imports
46.8%
48.9%
61.7%
59.2%
64.3%
61.7%
Current account balance as % of GDP
-11.9%
-10.5%
-4.4%
-2.9%
-1.2%
-3.8%
Budget deficit as % of GDP
-9.0%
-9.6%
-13.9%
-9.3%
-8.2%
-8.2%
Public investment growth
7.5%
4.9%
9.0%
3.0%
-6.0%
9.8%
Money growth
26.7%
65.8%
113.3%
86.3%
45.7%
22.6%
Inflation
12.8%
19.3%
86.1%
128.4%
23.2%
8.4%
Exchange rate depreciation
14.6%
183.3%
62.2%
77.8%
7.1%
16.6%
Growth of real GDP
6.9%
6.9%
4.0%
4.0%
5.7%
5.7%
Aid as % of GDP
18.0% 17.5%
Foreign direct investment as % of GDP
8.6%
4.4%
Source: ADB (various issues), UN (various issues), IMF (various issues).
14.2%
2.3%
15.2%
3.9%
13.2%
4.6%
13.9%
5.2%
Table A7.1.2: Simulation results of counterfactual analysis: 1995-2001
Year
1995
1996
1997
1998
1999
2000
2001
GY
7.0
6.9
6.9
4.0
4.0
5.7
5.7
GYB
8.2
8.2
7.6
4.67
4.67
6.37
6.37
GYEX1
GYEX2
8.2
8.2
8.2
8.2
7.6
7.6
4.67
4.54
4.57
5.61
7.01
8.11
8.01
9.00
CAY
-12.4
-17.6
-21.4
-10.6
-10.7
-10.0
-11.5
CAYB
-15.6
-18.0
-16.1
-14.2
-14.5
-14.7
-14.9
CAYEX1
CAYEX2
-15.6
-15.6
-18.0
-18.0
-16.1
-16.1
-14.2
-14.2
-14.5
-14.3
-14.4
-14.1
-14.4
-14.1
Legend: GY is output growth, CAY is current account balance, B is baseline model, EX1 is simulation result of
experiment 1, and EX2 is simulation result of experiment 2.
Chapter 7 __________________________________________________________________Page 181
Appendix 7.2: Data used in the counterfactual analysis
This appendix consists of three tables. Table A7.2.1 presents data used to
create the baseline model. Table A7.2.2 presents data used for the policy
experiment to disentangle the effect that the improvement in donors’
conditionality may have on output growth and the current account deficit.
Table A7.2.3 presents data used for the policy experiment to disentangle the
effect that the improvement in the state and institutional capability of the Lao
PDR may have on output growth and the current account deficit.
Table A7.2.1: Data used for creating the baseline model
Year
IGY
GLE
DWEA
YF
MCY
FDIY
REER
INF
1998
15.5
-61.5
0
1
12
2.3
80
12.8
1999
14.4
-73.4
0
1
12
3.9
80
71.2
2000
12.8
-30.8
0
1
12
4.6
80
120.4
2001
13.3
-5.7
0
1
12
5.2
80
36.1
Legend: IGY is public investment, GLE is the growth rate of effective labour force, DWEA is dummy variable YF is the
weighted average by trade share of growth in per capita income of the Lao PDR’s trading partners, MCY is imports of
consumption goods, FDIY is total foreign direct investment, REER is the real effective exchange rate, INF is inflation
rate.
Table A7.2.2: Data used for experiment 1
Year
IGY
GLE
DWEA
YF
MCY
FDIY
REER
INF
1998
15.5
-63.5
0
1
12
2.3
80
12.8
1999
14.4
-56.8
0
1
12
3.9
100
12
2000
12.8
-12.3
0
1
12
4.6
100
12
2001
13.3
-5.8
0
1
12
5.2
100
12
Legend: IGY is public investment, GLE is the growth rate of effective labour force, DWEA is dummy variable YF is the
weighted average by trade share of growth in per capita income of the Lao PDR’s trading partners, MCY is imports of
consumption goods, FDIY is total foreign direct investment, REER is the real effective exchange rate, INF is inflation
rate.
Table A7.2.3: Data used for experiment 2
Year
IGY
GLE
DWEA
YF
MCY
FDIY
REER
INF
1998
15.5
-63.5
0
1
12
4.8
80
12.8
1999
14.4
-56.8
0
1
12
5.3
100
12
2000
12.8
-12.3
0
1
12
5.9
100
12
2001
13.3
-5.8
0
1
12
6.4
100
12
Legend: IGY is public investment, GLE is the growth rate of effective labour force, DWEA is dummy variable YF is the
weighted average by trade share of growth in per capita income of the Lao PDR’s trading partners, MCY is imports of
consumption goods, FDIY is total foreign direct investment, REER is the real effective exchange rate, INF is inflation
rate.
Chapter 8__________________________________________________________________ Page 182
Chapter 8
SUMMARY AND FURTHER RESEARCH
The evidence over the long term shows that when per capita
income increases in LDCs, there is strong domestic savings
effort… if growth can be sustained, therefore, significant
increases in domestic resource mobilization may be expected
which would, in due time, reduce dependence on external
finance and usher in the possibility of a more self-sustained
growth process.
UNCTAD (2000a, p. iii)
…States can improve their capabilities by reinvigorating their
institutions. This means not only building administrative or
technical capacity but instituting rules and norms that provide
officials with incentives to act in the collective interest while
restraining arbitrary action and corruption. An independent
judiciary, institutional checks and balances through the
separation of powers, and effective watchdogs can all restrain
arbitrary state action and corruption. Competitive wages for civil
servants can attract more talented people and increase
professionalism and integrity…
Stiglitz (1998, p. 29)
8.1 Introduction
The belief that foreign aid promotes sustainable economic growth in
developing countries has permeated much thinking in international
development co-operation throughout the past decades. Particularly since
the 1980s, aid donors have attached policy conditionality to aid that aimed to
support the aid-receiving country to achieve economic growth with external
viability. The quests for the effectiveness of aid have been focused on the
link between economic growth and the internal and external balances, as
explained in the two-gap model. A number of aid effectiveness studies have
also incorporated variables for policy indices, institutional quality and
political instability into the analytical framework in order to ascertain the
effectiveness of aid. However, there remains no clear-cut explanation as to
Chapter 8__________________________________________________________________ Page 183
why policy conditionality attached to aid might not always promote
sustainable economic growth in many developing countries.
To a great extent the perceived lack of evidence on the effectiveness of aid
has generally been blurred by an incomplete theory of the aid-growth nexus,
and by the associated econometric methodologies applied. As such, this
study has developed a theoretical model of the aid-growth nexus to include
the problems of policy conditions designed for aid delivery and policy
mismanagement in the recipient country caused by the lack of state and
institutional capability. The aid-growth theory has been empirically tested for
the case of the Lao PDR using two macroeconometric models. The
hypothesis relates to whether policy conditionality attached to aid does or
does not help the recipient country to raise investment with external viability.
A summary of the contributions to knowledge on the effects of foreign aid on
economic growth, and the implications for policy makers are provided in
Section 8.2. Some suggestions for further research in this area are
discussed in Section 8.3.
8.2 Summary and implications
The discussions in Chapter 1 indicate that sound economic management
and the quality of state and institutional capability in the recipient country do
matter for the improvement of aid effectiveness. Chapter 2 has clarified this
point by surveying the literature on aid effectiveness. It is obvious from this
survey that there remains no clear-cut explanation for the interactive effect of
aid and policy conditionality on economic growth in two strands of aid
effectiveness studies. The first strand of aid effectiveness studies focused on
the effects of aid on economic growth within the framework of growth
theories. Many aid effectiveness studies have drawn on a flourishing
empirical growth literature to employ sophisticated models to address some
of the theoretical and econometric weaknesses of earlier aid effectiveness
studies. In these aid effectiveness studies, attention has been paid to the
changes in recipient country’s policies and institutions that may influence
Chapter 8__________________________________________________________________ Page 184
economic growth. The studies also regard political stability, a stable
macroeconomic environment, an open trade regime, protection of property
rights and the improvement in the quality of public services as factors
necessary to maintain sustainable economic growth. However, the studies
have failed to explain why these factors had more positive outcomes in some
countries than others, and some circumstances rather than others. The
studies have also failed to explain why these same positive outcomes may or
may not be sustainable. In other words, little attention has been paid to the
heterogeneous nature of the developing world. The presumption that all
economies are similar and thus will respond in a similar manner to the same
policy changes does not always hold.
The second strand of aid effectiveness studies has focused on the
possibilities that higher aid inflows may not always raise investment, as
assumed in the “two-gap” theory. One possibility is that an increase in aid
inflow may displace domestic savings and/or “crowd out” private investment.
This circumstance could be related to the behaviour of the recipient
government in conducting fiscal policy to accommodate aid inflow. In other
words, if an increase in aid inflow is used to invest in low productivity sectors
and/or to increase government consumption spending and/or to reduce tax
revenue, that would make aid less effective and discourage domestic
savings and investment. This is due to the issue of aid fungibility. Another
possibility is that foreign aid may impair economic growth via the impact of a
high level of aid inflow on the real effective exchange rate. This impact could
erode export earnings, which in turn reduces the ability to import capital
goods. Thus, the recipient country might be unable to increase investment
as required. This is also the well-known issue of the “Dutch disease”.
However, the aid effectiveness studies have failed to question the quality of
policy design to mitigate the side effects arising from the crowding-out effect
of fiscal policy and the issue of “Dutch Disease”. In fact, the set of policy
prescriptions propagated through the “adjustment programmes” has
overlooked the underdeveloped nature of the Least Developed Countries
Chapter 8__________________________________________________________________ Page 185
(LDCs) economies. Essentially, this set of policy prescriptions has neglected
the structural constraints faced by many LDCs, such as an extremely low
level of domestic savings, a lack of legal and regulatory frameworks, a
vulnerable indigenous private sector, a lack of market access for their export
commodities and a high degree of vulnerability to external shocks.
Therefore, it was often the case that policy prescriptions under the
Washington Consensus could only solve parts of the problems facing the
LDCs’ development efforts.
To clarify the aforementioned problems of aid effectiveness, a theoretical
model of the aid-growth nexus has been developed in Chapter 3. The
analytical framework has been derived from the development issues facing
LDCs. The problems of policy conditionality and policy mismanagement in
the recipient country have been included in the analytical framework. It has
been argued that the interruption of the “adjustment programmes” often
relates to inadequate policy design (i.e., lack of contingency measures to
deal with external shocks, and tensions between policy conditionality and the
recipient country’s national sovereignty) and the problems caused by the
lack of state and institutional capability (i.e., lack of enforcement of the rule
of law and problems of policy mismanagement).
The interruption of the “adjustment programmes” in turn leads to the
instability of aid flows. Two testable propositions have been demonstrated to
explain why policy conditionality attached to aid might not always promote
sustainable economic growth in the LDCs. First, the model has simulated
that a stable aid flow contributes to economic growth even when aid is
fungible. Second, the model has also simulated that unstable aid inflow
impairs the favourable effect of stable aid inflow. The theoretical analysis
undertaken here makes a contribution to the knowledge of aid effectiveness
in the LDCs. It is suggested that the contribution of aid to economic growth
depends not only on the ability of aid to increase investment in the recipient
country but also the quality of policy conditionality and the state and
Chapter 8__________________________________________________________________ Page 186
institutional capability of the recipient country to implement policy
conditionality.
To
test
the
hypothesis
established
above,
in
Chapter
4
two
macroeconometric models have been developed for a case of the Lao PDR.
These macroeconometric models are developed to test the economy-wide
effects of aid resulting from the imposition of policy conditionality by aid
donors. This analysis therefore departs from the traditional aid fungibility and
aid-growth models. The macroeconometric technique has been utilized to
solve the methodological weakness in the aid fungibility and aid-growth
models makes a contribution to knowledge. As the conventional model of aid
fungibility
confines
itself
only
to
the
government
sector,
the
macroeconometric model of aid fungibility model developed here allows aid
inflows to have direct, indirect and feedback effects on public investment,
private investment and external trade. As applied to the case of the Lao
PDR, the effects of stable aid inflows on economic growth have been
assessed through the potential side effects of aid fungibility on domestic
investment. The Lao Government’s fiscal behaviours in response to aid
inflows have been captured through the coefficient of aid fungibility.
Regarding the macroeconometric model of the aid-growth nexus, the
standard neo-classical growth model has been modified to integrate several
aspects of the potential effects of capital flows on the Lao PDR’s economy,
as explained by the two-gap model. The contribution here for the aid-growth
nexus is that the model is analysed to track the dynamic impact of capital
inflows through the direct and indirect effects of capital inflow on the
investment and trade sectors. The dynamic multiplier analysis is employed to
measure the interactive effect of aid and policy conditionality on economic
growth. In this context, the aid-growth coefficient implicitly represents the
interaction between aid and the state and institutional capability and the
effect of this interaction on the Lao PDR’s economic growth. This technique
avoids the weakness of including policy and institutional indices as generally
encountered in the traditional aid-growth regression.
Chapter 8__________________________________________________________________ Page 187
The macroeconometric model of aid fungibility employed in Chapter 5
addresses the issue of selective strategies for aid allocation in the case of
the Lao PDR. The analysis in this Chapter provides an empirical contribution
to knowledge of aid effectiveness literature. The empirical finding in Chapter
5 reveals that aid was fungible but it did not crowd out private investment in
the Lao PDR. In displacement theory, the existence of aid fungibility has
always been attributed to malicious intent by the aid-receiving country.
However, the empirical finding in this chapter implies that aid fungibility is
associated with the lack of institutional capacity and the extreme poverty of
the country. For example, the observed decline in tax revenue was rooted in
the country’s extremely low tax base and the lack of fiscal regulatory
framework to ensure the efficiency of tax collection. From this viewpoint, it
would be wrong for the donors to use the aid fungibility issue to exclude poor
countries from receiving aid. As for the effect of aid fungibility on investment,
the indication that aid fungibility did not crowd out private investment may be
evidence of the improvement in fiscal discipline. In fact, as policy
conditionality attached to aid often requires the recipient country to reduce
its domestic borrowing, this leaves more resources available for increased
investment by the private sector. Therefore, imposition of the donors’
conditionality on fiscal policy reforms makes an increase in aid inflow
complementary to, and reduces the potential side effects of, aid fungibility on
private investment.
Given that a high level of stable aid inflow did not crowd out private
investment, a macroeconometric model of the aid-growth nexus has been
employed in Chapter 6 to further explore the effects of stable aid flow on
economic growth. The analysis carried out in this Chapter involves an
empirical contribution to knowledge of the aid effectiveness literature. The
finding indicates that aid does contribute to economic growth through public
investment. The effect of the “adjustment programmes” on economic growth
also reflects the trade-off between short-run stabilisation measures and longrun structural adjustment measures to support economic growth. In other
Chapter 8__________________________________________________________________ Page 188
words, the stabilisation measures, including budget cuts and inflation
reduction, are found to have a negative effect on private income and social
spending. Thus, stabilisation programmes retarded economic growth in the
short run. However, the success in restraining the demand side of the
economy at the outset of the “adjustment programmes” in the early 1990s
brought about stability in the macroeconomic environment. This favourable
environment enabled the Lao PDR to make progress in implementing the
structural adjustment programmes during the 1990-97 period. A large
amount of stable aid and FDI inflows in this period released the foreign
exchange constraints on capital imports, which in turn contributed to strong
growth in the export sector and enabled the Lao PDR to achieve higher
economic growth with external viability.
The analysis carried out in Chapters 5 and 6 reveals that stable aid inflows
contribute to economic growth even when aid is fungible. This implies that
the success in implementing policy conditionality along with stable aid flows
improves the effectiveness of aid. However, whether this outcome is
sustainable relies not only on improved and appropriate policy conditionality
but also on building good governance through strengthening state and
institutional capacity. Therefore Chapter 7 has evaluated the role of state
and institutional capability (i.e. good governance) in implementing policy
conditionality. The findings seem to suggest that lack of state and
institutional capacity creates many symptoms of development failures such
as poor quality public services, a lack of fiscal and monetary discipline, an
unsustainable current account deficit and rampant corruption. These
symptoms occur regardless of the political regime. These problems, coupled
with an inadequate policy design for aid delivery clearly reflect the
impossibility of rigorously enforcing policy conditions.
Chapter 7 also contributes to the knowledge of aid effectiveness by
demonstrating that while a stable aid flow has an overall favourable effect on
economic growth, the existence of temporary unstable aid flows impairs this
effect. To capture the effect of an unstable aid inflow on economic growth,
Chapter 8__________________________________________________________________ Page 189
counterfactual analysis is employed. In the case of the Lao PDR, the
interruption in the “adjustment programmes” was due to forward-looking
investment disagreements over the irrigation project, as well as a large
current account deficit at the time of the Asian financial crisis that led to the
withholding of financial aid by the IMF and other donors. Interruption in aid
flows has resulted in economic volatility, which in turn reduced investment
and led to a slow down in economic growth. This finding implies that policy
conditionality in the form of an “adjustment programmes” when attached to
aid can tend to conflict with its stated objective in promoting the stability of a
country’s macroeconomic environment and the sustainability of its economic
growth. This could be an explanation as to why policy conditionality attached
to aid might not always promote sustainable economic growth. To improve
the effectiveness of aid, it is necessary to ensure that the diagnosis for
policy formulation for each LDCs is right. This demands a constructive policy
dialogue between the donor and recipient countries. There is nothing wrong
in allowing the recipient country take the lead in the process of policy
formulation and the pace of policy implementation. In addition, promoting
national sovereignty, increasing aid coordination and urging the recipient
country to build good governance are certainly the right way to improve the
effects of international development cooperation in the 21st century.
8.3 Further research
It is important to recall that this study emphasised strategies for maintaining
economic growth with external viability which is at the heart of any
successful development for developing countries. Development success
depends on establishing a virtuous circle between investment, exports and
savings to ensure that a country is able to raise investment with external
viability.
It is obvious from this study that the Lao PDR’s ability to raise investment
remains heavily reliant on the stability of foreign capital inflows and
adequate reforms in the state and institutions. However, the analysis carried
Chapter 8__________________________________________________________________ Page 190
out in this study did not elaborate on ways to strengthen state and
institutional capability to ensure that the country could achieve sustainable
economic growth. Particularly, issues of maintaining the external viability are
often related to adopting a prudent macroeconomic policy, attracting foreign
aid and foreign direct investment and access to international trade. These
are some very complex issues still to be evaluated. How to advance the
capacity and capability of the state and institutions to improve the
macroeconomic environment deserves further study.
In addition, the main contribution of this study lies in the application of
macroeconometric modelling to analyse the potential effects of foreign aid
on economic growth. Although the models are capable of explaining how aid
inflow may influence economic performance in the Lao PDR, the model
developed in this study is in the aggregate form. It would be more interesting
if the impact of aid on economic growth can be desegregated in at least
three major sectors: (i) the agricultural sector, (ii) the industrial sector and
(iii) the services sector. By doing this, the channels though which aid may
affect economic growth could be highlighted.
Reference __________________________________________________________________Page 191
REFERENCES
ADB (1991) Key Indicators of Developing Asian and Pacific Countries,
Oxford: Oxford University Press.
ADB (1994) Key Indicators of Developing Asian and Pacific Countries,
Oxford: Oxford University Press.
ADB (1995) Governance: Sound Development Management, Manila: Asian
Development Bank.
ADB (1997) Key Indicators of Developing Asian and Pacific Countries,
Oxford: Oxford University Press.
ADB (1999) Asian Development Outlook 1999, Hong Kong ; Oxford : Oxford
University Press.
ADB (2001) Asian Development Outlook 2000, Hong Kong ; Oxford : Oxford
University Press.
ADB (2001) Key Indicators of Developing Asian and Pacific Countries,
Oxford: Oxford University Press.
Agenor, P. and Montiel, P. (1996) Development Macroeconomics, New
Jersey: Princeton University Press.
Alesina, A. and Perotti, R. (1995) “The Political Economy of Budget Deficits”,
IMF Staff Papers, Vol. 42, pp. 1-31.
Alesina, A. and Dollar, D. (1998) “Who Gives Aid to Whom and why?”,
NBER Working Paper 6612, National Bureau of Economic
Research, Cambridge, Mass.
Arndt, H. (1987) Economic Development: The History of An Idea, Chicago:
Chicago University Press.
Banerjee, A., Dolado, J., Galbraith, J., and Hendy, D. (1993) Co-integration,
Error Correction, and The Econometric Analysis of Nonstationary Data, New York: Oxford University Press.
Reference __________________________________________________________________Page 192
Bank of Thailand (1992) Economic and Financial Report of The Northeast,
Bangkok: The Bank of Thailand.
Bank of The Lao PDR (1993) Annual Report 1993, Vientiane: Pakpasack
Kanphim.
Bank of The Lao PDR (1994) Annual Report 1994, Vientiane: Pakpasack
Kanphim.
Bank of The Lao PDR (1995) Annual Report 1995, Vientiane: Pakpasack
Kanphim.
Bank of The Lao PDR (1996) Annual Report 1996, Vientiane: Pakpasack
Kanphim.
Barro, R. (1991) “Economic Growth in a Cross Section of Countries”,
Quarterly Journal of Economics, Vol. 106, pp. 407-443.
Barro, R. and Sala-i-Martin, X. (1995) Economic Growth, New York:
McGraw-Hill.
Beynon, J. (2001) “Policy Implications for Aid Allocations of Recent
Research on Aid Effectiveness and Selectivity”, the paper
was presented at the Joint Development Centre/Dac Experts
Seminar on "Aid Effectiveness, Selectivity and Poor
Performers", OECD, Paris, 17 January 2001, mimeo.
Available online at:
http://www.ercdfid.org.uk/knowledgecentre/JB-DAC2.pdf
Beynon, J. (1999) “A Critique of The World Bank Research”, mimeo.
Available on line at:
http://www.dfid.gov.uk/public/what/advisory/group1/beynon_
annex2.html
Blejer, M., Cheasty, A. (1989) “Fiscal Policy and Mobilization of Savings for
Growth”, in: Blejer, M.I., Chu, K.-Y. (Eds), Fiscal Policy,
Stabilization, and Growth in Developing Countries. IMF,
Washington, D.C.
Boone, P. (1994) The Impact of Foreign Aid on Savings and Growth, London
School of Economics (June), mimeo.
Reference __________________________________________________________________Page 193
Boone, P. (1996) “Politics and The Effectiveness of Foreign Aid”, European
Economic Review, Vol. 40, pp. 290-329.
Borner, S. and Paldam, M. (1998) The Political Dimension of Economic
Growth, New York: St. Martin’s Press.
Bourdet, Yves. (2000) The Economics of Transition in Laos : From Socialism
to ASEAN Integration, Northampton, MA : Edward Elgar.
Bourdet, Yves. (2001) “LAOS in 2000: The Economics of Political
Immobilism”, Asian Survey, Vol.XLI (1), pp. 164-170.
Bourdet, Yves. (1993) “Laos: Macroeconomic Stabilization Policies”, Paper
No. 11/93, Institute of Economic Research Discussion,
University of Lund, Sweden.
Bowen, J. (1998) Foreign Aid and Economic Growth : A Theoretical and
Empirical Investigation, Sydney : Ashgate.
Bredenkamp, H. and Schadler, S., eds. (1999) Economic Adjustment and
Reform in Low-income Countries, Studies by the Stuff of the
International Monetary Fund, IMF, Washington DC.
Burnside, C. and Dollar, D. (1997) “Aid, Policies, and Growth”, Policy
Research Working Paper 1777., World Bank, Development
Research Group, Washington, D.C.
Burnside, C. and Dollar, D. (1998) “Aid, The Incentive Regime, and Poverty
Reduction”, Policy Research Working Paper 1937, June
1998, World Bank, Development Research Group,
Washington, D.C.
Cashel-Cordo, P. and Craig, S. (1990) “The Public Sector Impact of
International Resource Transfers”, Journal of Development
Economics, Vol. 32, pp. 17-42.
Cassen, R. (1994) Does Aid Work, 2nd (edition), Oxford: Clarendon Press.
Challen,
D.
and Hagger, A. (1983) Macroeconometric Systems:
Construction, Validation and Application, London: Macmillan.
Reference __________________________________________________________________Page 194
Chenery, H. and Eckstein, P. (1970) “Development Alternatives for Latin
America”, Journal of Political Economy, Vol. 78, July/Aug.,
pp. 966-1006.
Chenery, H. and Strout, A. (1966) “Foreign Assistance and Economic
Development”, American Economic Review, Vol. 66(4), pp.
679-733.
Chow, G. (1975) Analysis and Control of Dynamic Economic Systems, New
York: Wiley.
Collier, P. and Dollar, D. (1999a) “Aid Allocation and Poverty Reduction”,
Policy Research Working Paper 2041, January 1999, World
Bank, Development Research Group, Washington, D.C.
Collier, P. and Dollar, D. (1999b) “Aid Allocation and Poverty Reduction”,
mimeo. Available online at:
http://www.dfid.gov.uk/public/what/advisory/group1/beynon_
annex1.html
Collier, P. and Gunning, W. (1992) “Aid and Exchange Rate Adjustment in
African Trade Liberalisation”, Economic Journal 102(413),
pp. 925-39.
Connors, T. (1979) “The Apparent Effects of Recent IMF Stabilization
Programs”, International Discussion Papers 135. Board of
Governors of the Federal Reserve System.
Conway, P. (1994) “IMF Lending Programs: Participation and Impact.”
Journal of Development Economics 45 (2), pp. 355-391.
Corden, W. and Nearly, J. (1982) “Booming Sector and Deindustrialization in
a Small Open Economy”, The Economic Journal, Vol. 92, pp.
825-848.
Dalgaard, C. and Hansen, H. (2000) “On Aid, Growth, and Good Policies”,
CREDIT Research Paper 00/17, Centre for Research in
Economic Development and International Trade, University
of Nottingham.
Reference __________________________________________________________________Page 195
Deaton, A. (1977) “Involuntary Saving Through Unanticipated Inflation”,
American Economic Review, Vol.67(12), pp. 899-910.
Dixit, A. and Pindyck, R. (1994) Investment Uncetainty, New Jersey:
Princeton University Press.
Dollar, D. and Kraay, A. (2000) “Growth Is Good for The Poor” World Bank,
Development Research Group, Washington, D.C.
Dollar, D. and Easterly, W. (1998) “The Search for The Key: Aid, Investment
and Policies in Africa”, World Bank, Development Research
Group, Washington, D.C.
Dornbusch, R. and Fisher, S. (1994) Macroeconomics, 6
McGraw-Hill.
th
, New York:
Dornbusch, R. and Reynoso, A. (1993) “Financial Factors in Economic
Development”, in R. Dornbusch (ed.), Policy Making in The
Open Economy: Concept and Case Studies in Economic
Performance, Oxford University Press: New York. pp. 64-90.
Dowling, J. and Hiemenz, V. (1982) “Aid, Savings and Growth in The Asian
Region”, quoted in White, H. (1992a) “The Macroeconomic
Impact of Development Aid: A Critical Survey”, Journal of
Development Studies, Vol. 28 (2), pp. 163-240.
Durbarry, R., Gemmell, N. and Greenaway, D. (1998) “New Evidence on The
Impact of Foreign Aid on Economic Growth”, CREDIT
Research Paper 98/8, University of Nottingham.
Easterly, W. (1999) “The Ghost of Financial Gap: Testing The Growth Model
Used in The International Financial Institutions”, Journal of
Development Economics, Vol. 60 (2), pp. 423-438.
Easterly, W. (2000) “The Lost Decades: Explaining Developing Countries”
Stagnation 1980-1998”, mimeo. Available online at:
http://www.worldbank.org/research
Easterly, W. Islam, R. and Stiglitz, J. (2000) "Shaken and Stirred: Explaining
Growth Volatility." Annual Bank Conference on Development
Reference __________________________________________________________________Page 196
Economics, mimeo. Available online at:
http://www.worldbank.org/research/growth/paauthor.htm
Easterly, W. and Levine, R. (2001) "It's Not Factor Accumulation: Stylized
Facts and Growth Models", World Bank Economic Review
(forthcoming).
Edwards, S. and Van Wijnbergen, S. (1989) “Disequilibrium and Structural
Adjustment” In H. Chennery and T.N. Srinivasan edition
Handbook of Development Economics Volume II,
Amsterdam: North-Holland.
Ekman, B. and Metell, K. (1993) “Foreign Aid and Fungibility: The Kenyan
Case”, quoted in White, H. (1998) Aid and Macroeconomic
Performance: Theory, Empirical Evidence and Four Country
Cases, Institute of Social Studies, The Hague.
Enders, W. (1995) Applied Econometric Times Series, New York: Wiley
Ericsson, N. and Irons, J. (1995) “The Lucas Critique in Practice: Theory
Without
Measurement”,
in
K.
Hoover
(ed.),
Macroeconometric: Developments, Tensions and Prospects,
London: Kluver Academic Publishers.
Fair, R. (1994) Testing Macroeconometric Models, Massachusetts: Harvard
University Press.
Feyzioglu, T., Vinaya, S. and Min, Z. (1996) “Foreign Aid’s Impact on Public
Spending”, Policy Research Working Paper No. 1610, World
Bank, Washington D.C.
Feyzioglu, T., Vinaya, S. and Min, Z. (1998) “A Panel Data Analysis of The
Fungibility of Foreign Aid”, World Bank Economic Review
12(1).
Franco-Rodriquez, S. (2000) “Recent Advances in Fiscal Response Models
with and Application to Costa Rica”, Journal of International
Development, Vol. 12:3, pp. 429-442.
Reference __________________________________________________________________Page 197
Franco-Rodriquez, S., McGillivray, M. and Morrissey, O. (1998) “Aid and The
Public Sector in Pakistan: Evidence with Endogenous Aid”,
World Development, Vol.26, pp.1241-1250.
Frank, A. (1969) Latin America: Underdevelopment or Revolution, New York:
Columbia University Press.
Gang, I. and Khan, H. (1991) “Foreign Aid, Taxes, and Public Investment”,
Journal of Development Economics, Vol. 34, pp. 355-369.
Gavin, M. and Hausmann, R. (1998) “Macroeconomic Volatility and
Economic Development”. In Borner, S. and Paldam, M.
edition The Political dimension of Economic Growth, London:
Macmillian.
Goldfeld, S. and Quandt, R. (1972) Nonlinear Methods in Econometrics,
Amsterdam: North Holland.
Goldstein, M. (2000) “IMF Structural Conditionality: How Much Is Too
Much?”, mimeo, revision of paper presented in NBER
Conference on “Economic and Financial Crises in Emerging
Market Economies”, Woodstock, Vermont 19-21 October
2000.
Gonamee, K., Girma, S. and Morrissey, O. (2002) “Aid and Growth in SubSaharan Africa: Accounting for Transmission Mechanisms”,
CREDIT Research Paper No 02/05, University of
Nottingham.
Gounder, R. (1995) Overseas Aid Motivation: The Economics of Australia’s
Bilateral Aid, Aldershot: Avebury.
Gounder, R. and Xayavong, V. (2001) “Globalization and The Island
Economies of the South Pacific”, Discussion Paper No.
2001/4 WIDER and United Nations University, Helsinki.
Gounder, R. (2002) “Political and Economic Freedom, Fiscal Policy and
Growth Nexus: Some Empirical Results for Fiji”,
Contemporary Economic Policy, Vol. 20, pp. 234-245.
Reference __________________________________________________________________Page 198
Greenaway, D. (1998) “Does Trade Policy Reform Improve Economic
Performance?”, Scottish Journal of Political Economy, Vol.
45, pp. XX.
Griffin, K. (1970) “Foreign Capital, Domestic Savings and Economic
Development”, Oxford Bulletin of Economics and Statistics,
Vol. XXXII (2), pp. 99-112.
Griffin, K. and Enos J. (1970) “Foreign Assistance: Objectives and
Consequences”, Economic Development and Cultural
Change, Vol. 18 (3), pp. 313-26.
Grinols, E. and J. Bhagwati (1976) “Foreign Capital, Savings and
Dependence”, Review of Economics and Statistics, Vol. 58,
pp. 416-24.
Gujarati, D. N. (1995) Basic Econometric, (3rd eds), New York: McGraw-Hill.
Gupta, K. and Lensink, R. (1995) “Foreign Aid and The Public Sector: A
Simulation Approach”, mimeo, paper presented at the ESRC
study group, London, May 1995.
Gupta, K. (1975) “Foreign Capital Inflows, Dependency Burden and Savings
Rates in Developing Countries: A Simultaneous Equation
Model”, Kyklos, Vol, 28, pp. 358-374.
Gupta, K. (1993) “Sectoral Fungibility in Foreign Aid: Evidence from India”,
quoted in White, H. (1998) Aid and Macroeconomic
Performance: Theory, empirical Evidence and Four Country
Cases, Institute of Social Studies, The Hague.
Gylfason, T. (1987) “Credit Policy and Economic Activity in Developing
Countries with IMF Stabilization Programs. Studies in
International Finance 60, Princeton.
Haber, S. (2000) Political Institutions and Economic Growth in Latin America:
Essays in Policy, History, and Political Economy, California:
Hoover Institution Press.
Reference __________________________________________________________________Page 199
Hadjimichael, M., Ghura, D., Muhleisen, M., Nord, R. and Ucer, E. (1995)
“Sub-Saharan Africa: Growth, Saving, and Investment, 198693”, IMF Occasional Paper No. 118, IMF, Washington D.C.
Haggard, S., Lafay, J., and Morrisson, C. (1995) The Political Feasibility of
Adjustment in Developing Countries, OECD, Paris.
Hansen, H. and Tarp, F. (1999) “Aid Effectiveness Deputed”, CREDIT
Research Paper No 99/10, University of Nottingham.
Hansen, H. and Tarp, F. (2000) “Aid and Growth Regressions”, CREDIT
Research Paper No 00/7, University of Nottingham.
Hargreaves, C. (1992) Macroeconomic Modelling of The Long Run,
Worcester: Billing.
Hayter, T. (1971) Aid as Imperialism, Harmondsworth: Penguin.
Hayter, T. (1989) Exploited Dearth: British Aid and The Environment,
London: Earthscan.
Hayter, T. and Watson, C. (1985) Aid: Rhetoric and Reality, London: Pluto.
Heller, P. (1975) “A Model of Public Fiscal Behaviour in Developing
Countries: Aid, Investment, and Taxation”, American
Economic Review, Vol. 65 (3), pp. 429-445.
Heller, P., Bovenber, A. Catsambas, T., Chu, L. and Shome, P. (1988) “The
Implications of Fund-Supported Programs for Poverty”, IMF
Occasional Paper 58, Washington, D.C.
Hendry, D. and Richard, J. (1983) “The Econometric Analysis of Economic
Time Series”, International Statistical Review, Vol. 51, pp. 333.
Hibbs, D. (1977) “Political Parties and Macroeconomic Policy”, American
Political Science Review, Vol.71 pp. 1467-1487.
Hoover,
K.
(1995) Macroeconometric: Development,
Prospects, Boston: Kluwer.
Tensions,
and
Reference __________________________________________________________________Page 200
Hossain, A. and Chowdhury, A. (1998) Open-Economy Macroeconomics for
Developing Countries, Massachusetts: Edward Elgar.
Hutchcroft, P. (1994) “Booty Capitalism: Business, Government Relations in
The Philippines”, in A. Maclntyre (ed), Business and
Government in Industrialising Asia, Sydney: Allen and Unwin.
Juster, T. and Watchel, P. (1972a) “Inflation and the Consumer”, Brooking
Paper:1, pp. 71-114.
Juster, T. and Watchel, P. (1972b) “A Note on Inflation and the Saving
Rate”, Brooking Paper:3, pp. 765-778.
IMF
(1996)
“Lao People’s Democratic Republic-Recent Economic
Developments” Staff Country Report, No. 96/54, IMF,
Washington, D.C.
IMF
(1997)
“Lao People’s Democratic Republic-Recent Economic
Developments” Staff Country Report, No. 97/63, IMF,
Washington, D.C.
IMF
(1998a)
“Lao People’s Democratic Republic-Recent Economic
Developments” Staff Country Report, No. 98/77, IMF,
Washington, D.C.
IMF (1998b) “IMF Concludes Article IV Consultation with Lao People’s
Democratic Republic”, Press Information Notice (PIN) No.
98/48, Washington, D.C.
IMF (1998c) External Evaluation of The ESAF, Reported by a Group of
Independent Expert, IMF, Washington D.C.
IMF (1999a) “The IMF’s Enhanced Structural Adjustment Facility (ESAF): Is
It Working?, mimeo. Available on-line at:
http://www.imf.org/external/pubs/ft/esaf/exr/
IMF (1999b) “IMF Concludes Article IV Consultation with Lao People’s
Democratic Republic”, Press Information Notice (PIN) No.
99/109, Washington, D.C.
Reference __________________________________________________________________Page 201
IMF
(2000)
“Lao People’s Democratic Republic-Recent Economic
Developments” Staff Country Report, No.3, IMF,
Washington, D.C.
IMF (2001) Structural Conditionality in IMF-Supported Programs, Prepared
by the Policy Development and Review Department, IMF,
Washington D.C.
IMF (2002) International Financial Statistics, Data on CD-ROM, Washington
D.C.
Islam, A. (1992) “Foreign Aid and Economic Growth: An Econometric Study
of Bangladesh”, Applied Economics, Vol. 24 (5), pp. 541544.
Iweala, N., Kwakwa, V., Beckwith, A. and Ahmed, Z. (1999) “Impact of Asia’s
Financial Crisis on Cambodia and The Lao PDR”, Finance
and Development, Vol.36 (3) pp.1-7. Available on-line at:
http://www.imf.org/external/pubs/ft/fandd/1999/09/okonjo.htm
Karakitsos, E. (1992) Macrosystems: The Dynamics of Economic Policy,
Oxford: Blackwell.
Khan, M. (1987) “Macroeconomic Adjustment in Developing Countries: A
Policy Perspective”. Research Observer. Vol. 2 (1), pp. 2341.
Khan, M. and Knight, M. (1985) “Fund-Supported Adjustment Programs and
Economic Growth”, IMF Occasional Paper 41, Washington,
D.C.
Khilji, N. and Zampelli, E. (1991) “The Fungibility of US Assistance to
Developing Countries and The Impact on Recipient
Expenditures: a Case Study of Pakistan” World
Development, 19 (8), pp.1095-1105.
Killick, T. (1998) “Donor conditionality and Policy Reform”. In Borner, S. and
Paldam, M. edition The Political dimension of Economic
Growth, London: Macmillian.
Reference __________________________________________________________________Page 202
Killick, T. (1995) IMF Programs in Developing Countries: Design and Impact,
London: Routledge.
Krueger, A. (1997) “Trade Policy and Economic Development: How We
Learn, American Economic Review, Vol. 87, pp. 1-22.
Krueger, A. (1998) “Whither The World Bank and The IMF?”, Journal of
Economic Literature, XXXVI (4), pp. 1983-2020.
Lamfalussy, A. (1963) The United Kingdom and The Six, London: Macmillan.
Lao Government (1997) “Social Economic Development and Investment
Requirements 1997-2000”, Vientiane.
Lao Government (2001) “Report Prepared for The Third United Nation
Conference on The Least Developed Countries”, Brussels,
14-20
May,
mimeo.
Available
on-line
at:
http://www.unctad.org/en/docs/aconf191cp29lao.en.pdf
Lao People’s Revolutionary Party (2001) “Political Report of The Executive
Committee of The Central Committee of The Lao People’s
Revolutionary Party”, Presented at the Seventh Party
Congress, Vientiane, March 2001, mimeo. Available On-line
at:
http://www.mfa.laogov.net/EnglishVersion/Main/English.htm
Lavy, V. and Sheffer. (1991) Foreign Aid and Economic Development in The
Middle East: Egypt, Syria and Jordan, New York and
London: Praeger.
Lensink, R. and Morrissey, O. (1999) “Uncertainty of Aid Inflows and The
Aid-Growth Relationship”, CREDIT Research Paper 99/16,
University of Nottingham.
Lensink, R. and Morrissey, O. (2000) “Aid Instability as a Measure of
Uncertainty and The Positive Impact of Aid on Growth”,
Journal of Development Studies, Vol. 36 (3), pp. 31-44.
Lensink, R. and White, H. (1999) “Is There and Aid Laffer Curve?”, CREDIT
Research Paper 99/6, University of Nottingham.
Reference __________________________________________________________________Page 203
Lensink, R. and White, H. (2000) “Assessing Aid: A Manifesto for Aid in The
21st Century? Oxford Development Studies, Vol. 28, No 1,
pp. 5-17.
Levine, R. and Renelt, D. (1991) “Cross-Country Studies of Growth and
Policy: Some Methodological, Conceptual and Statistical
Problems”, World Bank Working Paper No. 608, Washington
D.C.
Lewis, W. (1955) Theory of Economic Growth, London: Allen and Unwin.
Lloyd, T., Morrissey, O. and Osei, R. (2001) “Aid, Exports and Growth in
Ghana”, CREDIT Research Paper 01/01, University of
Nottingham.
Lucas, R. Jr., (1976) “Econometric Policy Evaluation: A Critique”, in Brunner,
K. and Meltzer, A. (eds.), The Phillips Curve and Labor
Markets. Carnegie-Roschester conference Series on Public
Policy, Vol.1. Amsterdam: North-Holland.
Lucas, R. Jr, (1988) “On The Mechanics of Development Planning”, Journal
of Monetary Economics, Vol. 22 (1) pp. 3-42.
Maizels, A. and Nissanke, M. (1984) “Motivations for Aid to Developing
Countries”, World Development, Vol. 12 (9), pp. 879-900.
Malgrange, P. and Muet, A. (1984) Contemporary Macroeconomic
Modelling, Padstow: T. Press.
Massell, B., Pearson, S., and Fitch, J. (1972) “Foreign Exchange and
Economic Development: An Empirical Study of Selected
Latin American Countries”, Review of Economics and
Statistics, pp. 208-212.
Mauro, P. (1997) "The Effects of Corruption on Growth, Investment, and
Government Expenditure: A Cross Country Analysis." In:
Corruption and The Global Economy, 83-107. Edited by
Kimberly A.. Washington: The Institute for International
Economics.
Reference __________________________________________________________________Page 204
Mbaku, J. (1993) “Foreign Aid and Economic Growth in Cameroon”, Applied
Economics, Vol. 25 (10), pp. 1309-1314.
McCombie, S. and Thirlwall, A. (1997) Economic Growth and The Balanceof-Payments Constraint, Basingstoke, Hampshire :
Macmillan; New York : St. Martin's Press.
McGillivray, M. and Morrissey, O. (eds) (1998) Liberalisation and
Adjustment, London. Macmillan.
McGillivray, M. and Morrissey, O. (2000) “Aid Fungibility in Assessing Aid:
Red Herring or True Concern?, Journal of International
Development, Vol.12:3, pp.413-428.
McGillivray, M. and Morrissey, O. (2001) “A Review of Evidence on The
Fiscal Effects of Aid”, CREDIT Research Paper 01/13,
University of Nottingham.
McGillivray, M. and Oczkowski, E. (1991) “Modelling The Allocation of
Australian Bilateral Aid: A Two-Part Sample Selection
Approach”, Economic Record, Vol. 67 (197), pp. 147-51.
McGuire, M. (1987) “Foreign Assistance, Investment and Defence: A
Methodological Study with and Application to Israel”,
Economic Development and Cultural Change, Vol. 35, pp.
847-873.
McKinlay, R. and Little, R. (1977) “A Foreign Policy Model of U.S. Bilateral
Aid Allocation”, World Politics, Vol. XXX (1), pp. 58-86.
McKinlay, R. and Little, R. (1978a) “The French Aid Relationship: A Foreign
Policy Model of Distribution of French Bilateral Aid, 19641970”, Development and Change, Vol.9 (3), pp. 459-78.
McKinlay, R. and Little, R. (1978b) “A Foreign Policy Model of The
Distribution of British Bilateral Aid, 1960-70”, British Journal
of Political Science, Vol. 8 (3), pp. 313-31.
McKinlay, R. and Little, R. (1979) “The US Aid Relationship: A Test of The
Recipient Need and The Donor Interest Model”, Political
Studies, Vol. XXVII (2), pp. 236-50.
Reference __________________________________________________________________Page 205
McKinnon, R. (1973) Money and Capital in Economic Development,
Washington D.C: Brookings Institution.
McKinnon, R. (1993) The Order of Economic Liberalization: Financial Control
in The Transition to a Market Economy, Baltimore: John
Hopkins University Press.
Mecagni, M. (1999) “The Causes of Program Interruptions”. In Bredenkamp,
H. and Schadler, S., eds. (1999). Economic Adjustment and
Reform in Low-income Countries, Studies by the Staff of the
International Monetary Fund, Washington DC: IMF.
Mitsuhashi, K. (1996) An Economic Analysis of Aid Fungibility: Japan’s
Official Development Assistance To Indonesia Master of
Applied Economics Thesis, Massey University, New
Zealand.
Montes, M. Romeo A. and Somak T. (1995) Macroeconomic Management in
Southeast Asia’s Transitional Economies, Asian and Pacific
Development Centre, East-West Centre, UNDP.
Montes, M. (1995) Domestic Resource Mobilization in The Lao PDR and
Vietnam, Working Paper, East-West-Center, Honolulu.
Mosley, P. (1980) “Aid, Savings and Growth Revisited”, Oxford Bulletin of
Economics and Statistics, Vol. 42 (2), pp. 79-95.
Mosley, P. (1987) Overseas Aid: Its Defence and Reform, Brighton:
Wheatsheaf.
Mosley, P., Hudson, J. and Horrell, S. (1987) “Aid, The Public Sector and
The Market in Less Developed Countries”, Economic
Journal, Vol. 97 (388), pp. 616-41.
Mosley, P., Hudson, J. and Horrell, S. (1992) “Aid, The Public Sector and
The Market in Less Developed Countries: A Return to The
Scene of The Crime”, Journal of International Development,
Vol. 4, (2), pp. 139-50.
Reference __________________________________________________________________Page 206
Mosley, P., Harrigan, J. and Toye, J. (1995) Aid and Power: The World Bank
and Policy-Based Lending, New York: Routledge.
MPDF (1997) A Survey of Medium and Large Private Companies in The Lao
PDR, Mekong Project Development Facility, Vientiane, The
Lao PDR.
Murshed, S. (2000) “Globalization, Marginalization and Development”,
Working Paper No 175, United Nation University, World
Institute for Development Economic Research, Helsinki,
Finland.
Nayyar, D. (1997) “Globalization: The Game, The Players and The Rules” in
SD Gupta (ed), The Political Economy of Globalization,
London: Kluwer.
Naim, M. (1999) “Fads and Fashion in Economic Reforms: Washington
Consensus or Washington Confusion?” Paper prepared for
the IMF Conference on Second Generation Reforms,
Washington, D.C., mimeo. Available on line at:
http://www.ucm.es/info/eid/pb/Naim99WC.htm
Nelson, M. and Singh, R. (1998) “Democracy, Economic Freedom, Fiscal
Policy and Growth in LDCs: A Fresh Look”, Economic
Development and Cultural Changes, Vol. 46, pp. 677-696.
Nordhaus, W. (1975) “The Political Business Cycle”, The Review of
Economic Studies, Vol. 42, pp.169-190.
North,
D.
(1990) Institutions, Institutional Change and Economic
Performance, Cambridge: Cambridge University Press.
North, D. (1997) “The Contribution of The New Institutional Economics to An
Understanding of The Transition Problem”, WIDER Annual
Lectures 1, Helsinki: Hakapaino.
Nsouli, S., Peter, C. and Andreas, G. (1992) “Striving for Currency
Convertibility in North Africa”, Financial and Development,
29(4), pp.44-47.
Reference __________________________________________________________________Page 207
Nurkse, R. (1953) Problems of Capital Formation in Underdeveloped
Countries, Oxford: Blackwell.
OECD (1985) Report, Development Cooperation: Efforts and Policies of The
Members of The Development Association Committee, Paris:
OECD.
OECD (1997) Geographical Distribution of Financial Flows to Developing
Countries, Data on CD-ROM, Organisation for Economic Cooperation and Development, Paris.
OECD (2001) Geographical Distribution of Financial Flows to Developing
Countries, Data on CD-ROM, Organisation for Economic Cooperation and Development, Paris.
Otani, I., and Pham, Do Chi. (1996) “The Lao People’s Democratic Republic:
Systemic Transformation and Adjustment”, Occasional Paper
No 137, May, IMF, Washington, D.C.
Over, A. (1975) “An Example of The Simultaneous Equations Problem: A
Note on Foreign Assistance: Objectives and Consequences”,
Economic Development and Cultural Change, Vol. 24, pp.
751-6.
Pack, H. and Pack, J. (1990) “Is Foreign Aid Fungible? The Case of
Indonesia”, Economic Journal, Vol. 100 (399), pp. 188-194.
Pack, H. and Pack, J. (1993) “Foreign Aid and The Question of Fungibility”,
Review of Economics and Statistics, Vol. 75 (2), pp. 258-265.
Papanek, G. (1972) “The Effects of Aid and Other Resource Transfers on
Savings and Growth in Less Developed Countries”,
Economic Journal, Vol.82 (327), pp. 934-50.
Papanek, G. (1973) “Aid, Foreign Private Investment, Saving and Growth in
Less Developed Countries”, Journal of Political Economy,
pp. 137-138
Pastor, M., (1987) “The Effects of IMF Programs in The Third World: Debate
and Evidence from Latin America”, World Development,
Vol.15, pp.365-391.
Reference __________________________________________________________________Page 208
Persson, T. and Tabellini, G. (1990) Macroeconomic Policy, Credibility, and
Politics. London: Harwood.
Persson, T. and Tabillini, G. (2000) “Political Institutions and Policy
Outcomes: What Are The Stylize Facts?”, Institute for
International Economic Studies at the University of
Stockholm.
Persson, T. and Svensson, O. (1989) “Why a Stubborn Conservative Would
Run a Deficit: Policy with Time-Inconsistent Preference”, The
Quarterly Journal of Economics, Vol.104, pp.325-345.
Pesaran, M., and Shin, Y. (1995) “An Autoregressive Distributed Lag
Modelling Approach to Cointegration Analysis” DAE Working
Paper No. 9514, Department of Applied Economics,
University of Cambridge.
Pesaran, M. and Wickens, M. (1995) Handbook of Applied Econometrics:
Macroeconomics, Oxford: Blackwell.
Pesaran, M., Shin, Y. and Smith, R. (1996) “Testing for The Existence of
Long-Run Relationship, DAE Working Paper No. 9622,
Department of Applied Economics, University of Cambridge.
Pesaran, M. and Pesaran, B. (1997) Working with Microfit 4: Interactive
Econometric Analysis, Oxford: Oxford University Press.
Pindyck, R. and Rubinfeld, D. (1991) Econometric Models and Economic
Forecasts, 3rd (edition), Singapore: McGraw-Hill.
Przeworski, A. and Vreeland, J. (2000) “The Effect of IMF Programs on
Economic Growth”, Journal of Development Economics, Vol.
62, pp.385-421.
Ram, R. (1987) “Exports and Economic Growth in Developing Countries:
Evidence from Time-Series and Cross-Section Data”,
Economic Development and Cultural Change, Vol. 36 (1),
pp. 51-72.
Reference __________________________________________________________________Page 209
Rana, P. and Dowling, J. (1988) “The Impact of Foreign Capital on Growth:
Evidences from Asian Developing Countries”, The
Developing Economics, Vol. 26, pp. 3-11.
Reichel, R. (1995) “Development Aid, Saving and Growth in The 1980s: A
Cross-Section Analysis”, Saving and Development, Vol. 19
(3), pp. 52-73.
Romer, D. (1996) Advanced Macroeconomics, New York: McGraw-Hill.
Rostow, W. (ed) (1963) The Economics of Take-Off into Sustained Growth,
London: Macmillan.
Sachs, J., Botchwey, K., Cuchra, M. and Sievers, S. (1999). “Implementing
Debt Relief for The HIPCs”, Centre for International
Development, Harvard University.
Sala-i-Martin, X. (1997a) “I Just Ran Four Million Regression”, Colombia
University and Universitat Pompeu Fabra.
Sala-i-Martin, X. (1997b) “I Just Ran Two Million Regression”, American
Economic Review, Vol.87, pp.178-183.
SAS Institute (1993) SAS/ETS Software: Applications Guide 2: Econometric
Modelling Simulation, and Forecasting, Version 6, Cary, NC:
SAS Institute.
Savada, A. (1994) Laos: A Country Study, Washington: Library of Congress.
Available On-line at: http://lcweb2.loc.gov/frd/cs/latoc.html
Shaw, E. (1973) Financial Deepening in Economic Development, New York:
Oxford University Press.
Shi, M. and Svensson, J. (2000) “Conditional Political Business Cycles:
Theory and Evidence”, Harvard University and Institute for
International Economic Studies, Stockholm University.
Siermann, C. (1998) Politics, Institutions and The Economic Performance of
Nations, Northampton: Edwar Elgar.
Reference __________________________________________________________________Page 210
Sims, C. (1980) “Macroeconomics and Reality”, Econometrica, Vol.48, pp.148.
Sims, C. (1982) “Policy Analysis with Econometric Models”, Brookings
Papers on Economic Activity Vol 13(1), pp.107-52.
Snyder, D. (1990) “Foreign Aid and Domestic Savings: A Spurious
Correlation?”, Economic Development and Cultural Change,
Vol. 39 (1) (October), pp. 175-181.
Steger, T. (2000) “Economic Growth with Subsistence Consumption”,
Journal of Development Economics, Vol.62, pp. 343-361.
Stiglitz, J. (1998) “More Instruments and Broader Goals: Moving Toward The
Post-Washington Consensus”, WIDER Annual Lectures 2,
Helsinki: Hakapaino.
Stoneman, C. (1975), “Foreign Capital and Economic Growth”, Applied
Economics, Vol. 25, pp. 481-488.
Sun, H. (1998) Foreign Investment and Economic Development in China:
1979-1996, Sydney: Ashgate.
Swaroop, V., Jha, S. and Rajkumar, A. (2000) “Fiscal Effects of Foreign Aid
in a Federal System of Governance: The Case of India”,
Journal of Public Economics, Vol. 77, pp.307-30.
Thirlwall, A. (1980) Balance of Payments Theory and The United Kingdom
Experience, London: MacMillan.
Tsikata, T. (1998) “Aid Effectiveness: A Survey of The Recent Empirical
Literature”, Paper on Policy Analysis and Assessment, IMF,
Washington, D.C.
UN (1988) Statistical Yearbook for Asia and The Pacific, New York: United
Nations Publication.
UN (1991) Statistical Yearbook for Asia and The Pacific, New York: United
Nations Publication.
Reference __________________________________________________________________Page 211
UN (1994) Statistical Yearbook for Asia and The Pacific, New York: United
Nations Publication.
UN (1997) Statistical Yearbook for Asia and The Pacific, New York: United
Nations Publication.
UN (2000) Statistical Yearbook for Asia and The Pacific, New York: United
Nations Publication.
UNCTAD (1999) Trade and Development Report 1999, Geneva: UNCTAD.
UNCTAD (2000a) the Least Developed Countries 2000 Report, Geneva:
UNCTAD.
UNCTAD(2000b) Capital Flows and Growth in Africa, Geneva: UNCTAD.
Van Wijnbergen, S. (1986) Aid, Export Promotion and The Real Exchange
Rate: An African Dilemma, Macroeconomic division,
Development Research Department, World Bank, and
Centre for Economic Policy Research, London.
Voivodas, C. (1973) “Exports, Foreign Capital Inflow and Economic Growth”,
Journal of International Economics, Vol. 3, pp. 337-349.
Von Furstenberg (1980) “The Effect of Changing Size and Composition of
Government Purchase of Potential Output”, Review of
Economics and Statistics, Vol(62), No1.
Wallis, J. and Dollery, B. (2001) “Government Failure, Social Capital and
The Appropriateness of The New Zealand Model of Public
Sector
Reform
in
Developing
Countries”,
World
Development, Vol. 29 (2), pp. 245-263.
Weder, B. (1998) “Comment”, In Silvio Borner and Martin Paldam, The
Political Dimension of Economic Growth, (New York:
Macmillian and International Economic Association), pp.
351-356.
Weibsbrot, M., Naiman, R. and Kim, J. (2000) “The Emperor Has No Growth:
Declining Economic Growth Rates in The Era of
Reference __________________________________________________________________Page 212
Globalization” Internet Paper, Center for Economic and
Policy Research. Mimeo. Available on line at:
http://www.cepr.net/IMF/The_Emperor_Has_No_Growth.htm
Wescott, C. (2001) Key Governance Issues in Cambodia, The Lao PDR,
Thailand, and Vietnam, Manila: Asian Development Bank.
Weisman, E. (1990) “Aid as A Booming Sector: Evidence From a
Computable General Equilibrium Model of Papua New
Guinea”, Islands Australia Working Paper No. 90/13,
Canberra: National Centre for Development Studies, The
Australian National University.
Weisskopf, T. (1992) “Impact of Foreign Capital Inflows on Domestic
Savings in Under-Developed Countries”, Journal of
International Economics, Vol. 2, pp. 25-38.
White, H (1994) “Foreign Aid, Taxes and Public Investment: A Further
Comment”, Journal of Development Economics, Vol.45,
pp.155-163.
White, H. (1992a) “The Macroeconomic Impact of Development Aid: A
Critical Survey”, Journal of Development Studies, Vol. 28 (2),
pp. 163-240.
White, H. (1992b) “What Do We Know About Aid’s Macroeconomic Impact?
An Overview of The Aid Effectiveness Debate”, Journal of
International Development, Vol. 4 (2), pp. 121-37.
White, H. (1998) Aid and Macroeconomic Performance: Theory, empirical
Evidence and Four Country Cases, New York: St. Martin’s
Press.
White, H. and McGillivray, M. (1992) “Aid, The Public Sector and Crowding
in”, Institute of Social Studies Working Paper No. 126, The
Hague.
Williamson, J. (1990) "What Washington Means by Policy Reform", in J.
Williamson, ed., Latin American Adjustment: How Much Has
Happened?
Washington:
Institute
for
International
Economics.
Reference __________________________________________________________________Page 213
World Bank (1989) World Development Report 1989, Washington D.C.:
World Bank.
World Bank (1994) Lao People’s Democratic Republic: country economic
memorandum, The World Bank, Washington, D.C.: World
Bank.
World Bank (1995) World Development Report 1995, Press, Washington
D.C.: Oxford University Press.
World Bank (1996) “Structural Adjustment in The Lao PDR”, Precis N0.110,
Operations Evaluation Department, World Bank, mimeo.
Available On-line at:
http://wbln0018.worldbank.org/oed/oeddoclib.nsf/e90210f18
4a4481b85256885007b1724/8f126b7f9d02a673852567f500
5d8bd5?OpenDocument
World Bank (1997) World Development Report 1997: The State in Changing
World, New York: Oxford University Press.
World Bank (1998) Assessing Aid, Washington D.C.: Oxford University
Press.
World Bank (1999) World Development Indicators, Data on CD-ROM,
Washington D.C.
World Bank (2000) “The Lao PDR at a Glance”, mimeo. Available online at:
http://www.worldbank.org
World Bank (2001) World Development Report 2001: Building Institutions for
Markets, New York: Oxford University Press.
Reference __________________________________________________________________ Page 1